{ "schemaVersion": 1, "publicationId": "2026-10-19-windows-small-model-core", "sourceReportSha256": "b30dd40c3aeb6715502f24cc84ad22b0a7fe5697134a697455467c5d6e4a6aa1", "sourceReport": "evals/runs/small-model-core-2026-10-19/final-comparison.json", "sourceGeneratedAt": "2026-10-22T20:22:54.011Z", "complete": true, "newInferenceCalls": 0, "totals": { "trials": 264, "passed": 182, "modelFailures": 70, "infrastructureExclusions": 12, "artifactReviews": 254, "noArtifact": 10, "eligibleReviews": 247 }, "models": [ { "model": "gemma4-12b-q4", "n": 33, "pass": 27, "modelFailure": 6, "infra": 0, "reviewed": 33, "noArtifact": 2, "advisoryEligible": 31 }, { "model": "gemma4-e4b-q4", "n": 33, "pass": 25, "modelFailure": 8, "infra": 0, "reviewed": 33, "noArtifact": 0, "advisoryEligible": 33 }, { "model": "qwen3.5-9b-q4", "n": 33, "pass": 24, "modelFailure": 9, "infra": 0, "reviewed": 33, "noArtifact": 0, "advisoryEligible": 33 }, { "model": "nanbeige4.2-3b-q4", "n": 33, "pass": 12, "modelFailure": 9, "infra": 12, "reviewed": 33, "noArtifact": 6, "advisoryEligible": 21 }, { "model": "ornith1.5-9b-q4", "n": 33, "pass": 21, "modelFailure": 12, "infra": 0, "reviewed": 33, "noArtifact": 2, "advisoryEligible": 31 }, { "model": "qwen3.5-4b-q4", "n": 33, "pass": 18, "modelFailure": 15, "infra": 0, "reviewed": 33, "noArtifact": 0, "advisoryEligible": 33 }, { "model": "underdog-saluki-27b-iq2-mix", "n": 33, "pass": 29, "modelFailure": 4, "infra": 0, "reviewed": 33, "noArtifact": 0, "advisoryEligible": 32 }, { "model": "qwen3.8-27b-q2", "n": 33, "pass": 26, "modelFailure": 7, "infra": 0, "reviewed": 33, "noArtifact": 0, "advisoryEligible": 33 } ], "modelMetadata": { "gemma4-12b-q4": { "label": "Gemma 4 (12B, Q4)", "parameterSize": "12B", "quantization": "UD-Q4_K_XL", "tier": "medium", "modelManifestSha256": "1c9ef1ad0239eda01dc7194e49ee9e41b9a62e426a6e328ef1e384ce95e13f3a", "catalogManifestSha256": "fa33c9bd951804f7e3e6fc808b44b5243dcccd86ebb12c7ad6d17dadf79d4917", "weightsSha256": { "gemma-4-12B-it-qat-UD-Q4_K_XL.gguf": "90fd44e29e0d7cffeb0fd00dc73cfdab9ed0b0e95306ecf7821ea634c940c370", "mtp-gemma-4-12B-it-Q4_0.gguf": "fcb35dea42c71333db904cee11baac525c9ef872818ee3753f6cb156f3c6f4f6" } }, "gemma4-e4b-q4": { "label": "Gemma 4 (E4B, Q4)", "parameterSize": "8B", "quantization": "UD-Q4_K_XL", "tier": "small", "modelManifestSha256": "28a53d56f2de1801139624b48fedea7a66e418bd2e06936943b998647b4f507b", "catalogManifestSha256": "1345cf4383e82884bb0117208280dd4dfcb5626b154885eedc8a495715ea92fa", "weightsSha256": { "gemma-4-E4B-it-qat-UD-Q4_K_XL.gguf": "df0fd4ee07072c607c29a0a1cb4f98918426cca12f45a2776bdd6ee6d09a4de3", "mtp-gemma-4-E4B-it-Q4_0.gguf": "423074e537504b4f9ec5eafed5c639fac82c96631626efccacdd3c4039b20605" } }, "qwen3.5-9b-q4": { "label": "Qwen 3.5 (9B, Q4)", "parameterSize": "9B", "quantization": "Q4_K_M", "tier": "small", "modelManifestSha256": "0b6fc0d143e07c4f7681d07e3cec103150cb826ec585f77f327a0f558ec703f7", "catalogManifestSha256": "18901de3a7ad81ffad5f218df2d77a023072ce946d1422ec278ca44596a538db", "weightsSha256": { "Qwen3.5-9B-Q4_K_M.gguf": "e8dd94817e95d6c0939102049d068418269978377b13616c4726235e232841fe" } }, "nanbeige4.2-3b-q4": { "label": "Nanbeige 4.2 (3B, Q4)", "parameterSize": "3B", "quantization": "Q4_K_M", "tier": "tiny", "modelManifestSha256": "fbcfda7a3f3b0b648398c758c72ed9d2eeaa37e44c26afacd530de6b2087e4fd", "catalogManifestSha256": "7919e11bf440acb62cbd56fc828dddad23ed2a25ddbab7ec2325f545ff683b8f", "weightsSha256": { "Nanbeige_Nanbeige4.2-3B-Q4_K_M.gguf": "b92d2e35c9876b0c4bc671996360204f4149d0546c23d5954d5eb3b106c81f24" } }, "ornith1.5-9b-q4": { "label": "Ornith 1.5 (9B, Q4)", "parameterSize": "9B", "quantization": "Q4_K_M", "tier": "small", "modelManifestSha256": "5b35a560bfbed6f8490c9cda9beb47d4b4871f286e8b528769dc96710b1cfafa", "catalogManifestSha256": "9a921805fd83709c6607580e84d53e8429a5eae065d123dbd62abc340e13b219", "weightsSha256": { "Ornith-1.5-9B-Q4_K_M.gguf": "7d791afcb31812acc88cd5aafc675391df28c6fc3d8eae002bb4e6cc3d8cfd8d" } }, "qwen3.5-4b-q4": { "label": "Qwen 3.5 (4B, Q4)", "parameterSize": "4B", "quantization": "Q4_K_M", "tier": "tiny", "modelManifestSha256": "c0bb48afe269589f07943e7df6b34db47055590521c1b4107db33f5eb58df876", "catalogManifestSha256": "91624dbdad3774e034ba00a0e463b681038c49dc0c2fbca29836717e4bbe9ac9", "weightsSha256": { "Qwen3.5-4B-Q4_K_M.gguf": "3874209241c9a397e2f62cd3f70f80fd2dfbf0dfccb6838416bdb48a714e8630" } }, "underdog-saluki-27b-iq2-mix": { "label": "Underdog Saluki 1.0 (27B, IQ2 mix)", "parameterSize": "27B", "quantization": "IQ2-mix", "tier": "medium", "modelManifestSha256": "9d1e9b60b9da06be24dd18af6f298050005d2d7b0f760932838cbb93c7aeeea0", "catalogManifestSha256": "4499bf2701a67c86df31ee042fcd271720032bc15f0624f4217db528833885b0", "weightsSha256": { "Underdog-Saluki-27B-1.0-IQ2-mix.gguf": "4a673518b11b1c4445f9b9a9d3c40356f5ba39f6a6260dcb25f3fbfd475d9efb" } }, "qwen3.8-27b-q2": { "label": "Qwen 3.8 (27B, Q2)", "parameterSize": "27B", "quantization": "UD-Q2_K_XL", "tier": "medium", "modelManifestSha256": "4ef44841b637a583e7fc6862bd35ce54420158fb4cffde815cb420ce6ff61af5", "catalogManifestSha256": "4692cf9beea5dc4206184db694352b50327946cbcf3985266f30be9e8147031d", "weightsSha256": { "mmproj-F16.gguf": "cbb841a9ee0636b2ec172f5bb8df2ea8dfeb01e90fe7c6126581d662a0b4e43e", "Qwen3.8-27B-UD-Q2_K_XL.gguf": "fd4730dd8aad070517978752b63d530aeb1740d2283cab9fa24f1e404032ddb0" } } }, "sources": [ { "journal": "A", "root": "evals/runs/small-model-core-2026-10-19", "authoritativeTrials": 49, "summarySha256": "fede65c619aa44144560dfaa600aa6cbcce53eba31dd22d57d2b9a6931c26a7a", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "399efbfea8a2d8affca9636e303dba9d6f3a514696e27b40d4222718fae24275", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "b2334e47f2f5ca661a8ccd889c92f5d88c03f3138b677aea3b4dc93776677bbc" }, { "journal": "B", "root": "evals/runs/small-model-core-continuation-2026-10-20", "authoritativeTrials": 182, "summarySha256": "7a0b4dbac891c0597e2b9f8327cc7621cac1fdee45cd31f44ba98453e80b298e", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "6504c723b95e7c3d6ea46cf06d4dd4c821fe98634ab1b8e32fda76ea07fb924c" }, { "journal": "N1", "root": "evals/runs/nanbeige-32k-core-2026-10-20", "authoritativeTrials": 3, "summarySha256": "cb50a5a27fc86e4c1787788ccee6daa07df663bfa7d866836f2df45e8e708896", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "797cda69edd984df9903c6ae63601a91eebb3689362776aff14fc785ff2c82e6" }, { "journal": "N2", "root": "evals/runs/nanbeige-32k-core-continuation-2026-10-20", "authoritativeTrials": 2, "summarySha256": "25da0ce6641db5796e5353da7cf2739c61b61a2e7dead1f3d520097aac510203", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "b7c264c886f4746c3ca68c7a660cf4d9b132d639abf2b8366f640629dee413f4" }, { "journal": "N3", "root": "evals/runs/nanbeige-32k-core-continuation-2026-10-22", "authoritativeTrials": 2, "summarySha256": "3ce466511f24104ce10174fabcb506acf0ac0758c676da7a28ef72a68d67033a", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "f97c59c94d0d248893d933c48c62ee78a0fb7bc840f8a28bb049cb531d6fcb4c" }, { "journal": "N4", "root": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22", "authoritativeTrials": 1, "summarySha256": "89dd10f0df282e0178e916b2ae9e6de3374a778d60529d4d7aac9ea352fcd270", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "f3fb43b859be3db76a3b2e450a5a5ca84768e9d4370675965007699b2c79928a" }, { "journal": "N5", "root": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "authoritativeTrials": 6, "summarySha256": "69dd8746302bd7eed249a349351df09ee4fec64c9f4a06e79a4fd0e3611f71fa", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "9c9dc939ba5f407bcda0d06e4824f94349262e95c9f487f4d1a9fd88d45eee32" }, { "journal": "N6", "root": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22", "authoritativeTrials": 4, "summarySha256": "66bd4b1977b48e9634e1cd0973fac2f2a446e1b37a8eb7cdb111d65bc688484e", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "a864a14042f954248c3510add7989c1dcad5ed8988f114f950d4224b3d74e9a8" }, { "journal": "N7", "root": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22", "authoritativeTrials": 1, "summarySha256": "c353e3999ba2794421ef8a49dbbc7b9930036cd6f045bd713152b02ca9e37b69", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "b8b540b2c0cf97be2703c6f74554675180736a9f39cea5e129c383607ae23e96" }, { "journal": "N8", "root": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22", "authoritativeTrials": 3, "summarySha256": "16c0be73ab9830ef15c12530f0e32eafe11fe785d2feace4364147b52cadb1ad", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "6ea294ab1820d585e736908167a7a94545ab7c9b02bf6be832457b2318b8971f" }, { "journal": "N9", "root": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22", "authoritativeTrials": 2, "summarySha256": "0f3b83cdf378520981aaab92bf83dc8d3a8d79586507a2a2c2708c8018bff339", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "a7e63f7466f01bf2829983ba640cfd223847bc01372a2403b05d978120b5a27b" }, { "journal": "N10", "root": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22", "authoritativeTrials": 5, "summarySha256": "20738e6229ff56f0ce6764e4ff8c1777431abf3c9e09d369d08b55661a9d1b38", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "6565d69b5da59876c3e140df7afd2fb87100337491b39ca292e7fe562ff5b0f7" }, { "journal": "N11", "root": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22", "authoritativeTrials": 1, "summarySha256": "174b85e1ca6a38feee42d5792113eb52a9fce1c36ce2cc60137138b6922280d1", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "ac6f6913ee27b19b222e01f8252c877ac231ce43a0d0a28037e3e49d550ecd6d" }, { "journal": "N12", "root": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22", "authoritativeTrials": 3, "summarySha256": "acb065464465af65a0f68ca9e7f33c23e89bee3ebaca9747e96ce4d7e40028d4", "harnessCommit": "dee6f3e3aea8e40ee714038a7d805e34f0e1cbeb", "sourceHash": "98763c648f021292339838e264ab775024d6049bbfde0093d4f258df8cb1cde7", "serviceBuildHash": "a1ac336f2a669ac2e11ae528fe69e5bfaf91b5446a20364e3002de14aa6201f9", "runtimeHash": "c01c89e726cbc42cc5572a36a0de53db64a796fe9e48d4355dcda98aff8e5246", "catalogHash": "23ab3cfa5bde3d0753341e4572c7b3a8730debc9c171929344c49710a9a25784", "gildeVersion": "0.2.5", "driverHash": "0ed86ad3d15c7557789289d216c7437c1f2c43b2c2c1c808ed91daa573718383" } ], "rounds": [ { "id": "2026-10-20-rtx5070ti-core-original-64k", "journals": [ "A" ], "trials": 49 }, { "id": "2026-10-20-rtx5070ti-core-continuation-64k", "journals": [ "B" ], "trials": 182 }, { "id": "2026-10-20-rtx5070ti-nanbeige-core-32k", "journals": [ "N1", "N2", "N3", "N4", "N5", "N6", "N7", "N8", "N9", "N10", "N11", "N12" ], "trials": 33 } ], "additionalDiagnostics": { "oldNanbeige64K": 7, "interruptedQwen": 1, "includedInScorecard": false, "note": "Seven superseded Nanbeige 64K trials and the operator-interrupted Qwen attempt remain in the original archive, outside the 264 requested cells." }, "judging": { "provider": "codex-inline", "model": "current-session", "inputLimitBytes": 20480, "pending": 0, "artifactReviews": 254, "eligibleReviews": 247, "noArtifact": 10, "infrastructureDiagnosticReviews": 6, "protocolDeviationReviews": 1, "rubricAxes": { "default-html-v1": [ "visualQuality", "functionalCompleteness", "codeQuality", "polish" ], "campaign-code-v1": [ "functionalCompleteness", "codeQuality", "robustness", "scopeDiscipline" ], "campaign-data-v1": [ "dataAccuracy", "completeness", "traceability", "outputUsability" ], "campaign-doc-v1": [ "factualAccuracy", "functionalCompleteness", "actionability", "clarity" ] }, "aggregatesByJournalModelRubric": [ { "journal": "A", "model": "gemma4-12b-q4", "rubric": "default-html-v1", "n": 3, "mean": 4, "axes": { "visualQuality": 4.666666666666667, "functionalCompleteness": 3.6666666666666665, "codeQuality": 4.666666666666667, "polish": 3 } }, { "journal": "A", "model": "gemma4-12b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 4.5, "axes": { "functionalCompleteness": 4.666666666666667, "codeQuality": 4.666666666666667, "robustness": 3.6666666666666665, "scopeDiscipline": 5 } }, { "journal": "A", "model": "gemma4-12b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 3.25, "axes": { "dataAccuracy": 3, "completeness": 4, "traceability": 3, "outputUsability": 3 } }, { "journal": "A", "model": "gemma4-e4b-q4", "rubric": "default-html-v1", "n": 3, "mean": 4.25, "axes": { "visualQuality": 4.333333333333333, "functionalCompleteness": 5, "codeQuality": 5, "polish": 2.6666666666666665 } }, { "journal": "A", "model": "gemma4-e4b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 3.75, "axes": { "functionalCompleteness": 4, "codeQuality": 3.3333333333333335, "robustness": 3.3333333333333335, "scopeDiscipline": 4.333333333333333 } }, { "journal": "A", "model": "gemma4-e4b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 2.75, "axes": { "dataAccuracy": 3, "completeness": 3, "traceability": 3, "outputUsability": 2 } }, { "journal": "A", "model": "qwen3.5-9b-q4", "rubric": "default-html-v1", "n": 3, "mean": 3.5, "axes": { "visualQuality": 4, "functionalCompleteness": 3.6666666666666665, "codeQuality": 4, "polish": 2.3333333333333335 } }, { "journal": "A", "model": "qwen3.5-9b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 4.083333333333333, "axes": { "functionalCompleteness": 4.333333333333333, "codeQuality": 4, "robustness": 3.3333333333333335, "scopeDiscipline": 4.666666666666667 } }, { "journal": "A", "model": "qwen3.5-9b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 4.25, "axes": { "dataAccuracy": 4, "completeness": 5, "traceability": 5, "outputUsability": 3 } }, { "journal": "A", "model": "ornith1.5-9b-q4", "rubric": "default-html-v1", "n": 3, "mean": 4.416666666666667, "axes": { "visualQuality": 5.666666666666667, "functionalCompleteness": 4, "codeQuality": 4.666666666666667, "polish": 3.3333333333333335 } }, { "journal": "A", "model": "ornith1.5-9b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 4.333333333333333, "axes": { "functionalCompleteness": 4.333333333333333, "codeQuality": 4.666666666666667, "robustness": 3.3333333333333335, "scopeDiscipline": 5 } }, { "journal": "A", "model": "ornith1.5-9b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 2.75, "axes": { "dataAccuracy": 3, "completeness": 3, "traceability": 2, "outputUsability": 3 } }, { "journal": "A", "model": "qwen3.5-4b-q4", "rubric": "default-html-v1", "n": 3, "mean": 3.25, "axes": { "visualQuality": 4, "functionalCompleteness": 3, "codeQuality": 3.6666666666666665, "polish": 2.3333333333333335 } }, { "journal": "A", "model": "qwen3.5-4b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 3.25, "axes": { "functionalCompleteness": 2.6666666666666665, "codeQuality": 3.3333333333333335, "robustness": 2.6666666666666665, "scopeDiscipline": 4.333333333333333 } }, { "journal": "A", "model": "qwen3.5-4b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 3.25, "axes": { "dataAccuracy": 3, "completeness": 5, "traceability": 3, "outputUsability": 2 } }, { "journal": "A", "model": "underdog-saluki-27b-iq2-mix", "rubric": "default-html-v1", "n": 2, "mean": 4.25, "axes": { "visualQuality": 5, "functionalCompleteness": 4, "codeQuality": 4.5, "polish": 3.5 } }, { "journal": "A", "model": "underdog-saluki-27b-iq2-mix", "rubric": "campaign-code-v1", "n": 3, "mean": 4.666666666666667, "axes": { "functionalCompleteness": 4.666666666666667, "codeQuality": 5.333333333333333, "robustness": 3.6666666666666665, "scopeDiscipline": 5 } }, { "journal": "A", "model": "underdog-saluki-27b-iq2-mix", "rubric": "campaign-data-v1", "n": 1, "mean": 4.25, "axes": { "dataAccuracy": 4, "completeness": 5, "traceability": 4, "outputUsability": 4 } }, { "journal": "A", "model": "qwen3.8-27b-q2", "rubric": "default-html-v1", "n": 3, "mean": 4.5, "axes": { "visualQuality": 5.666666666666667, "functionalCompleteness": 3.6666666666666665, "codeQuality": 5, "polish": 3.6666666666666665 } }, { "journal": "A", "model": "qwen3.8-27b-q2", "rubric": "campaign-code-v1", "n": 3, "mean": 5.083333333333333, "axes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 4.333333333333333, "scopeDiscipline": 5 } }, { "journal": "A", "model": "qwen3.8-27b-q2", "rubric": "campaign-data-v1", "n": 1, "mean": 4.75, "axes": { "dataAccuracy": 5, "completeness": 5, "traceability": 4, "outputUsability": 5 } }, { "journal": "B", "model": "gemma4-12b-q4", "rubric": "default-html-v1", "n": 6, "mean": 3.5416666666666665, "axes": { "visualQuality": 3.6666666666666665, "functionalCompleteness": 3.8333333333333335, "codeQuality": 3.8333333333333335, "polish": 2.8333333333333335 } }, { "journal": "B", "model": "gemma4-12b-q4", "rubric": "campaign-code-v1", "n": 5, "mean": 4.5, "axes": { "functionalCompleteness": 4.6, "codeQuality": 5, "robustness": 3.4, "scopeDiscipline": 5 } }, { "journal": "B", "model": "gemma4-12b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 3, "axes": { "dataAccuracy": 3, "completeness": 4, "traceability": 3, "outputUsability": 2 } }, { "journal": "B", "model": "gemma4-12b-q4", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.479166666666667, "axes": { "factualAccuracy": 4.666666666666667, "functionalCompleteness": 4.416666666666667, "actionability": 4.25, "clarity": 4.583333333333333 } }, { "journal": "B", "model": "gemma4-e4b-q4", "rubric": "default-html-v1", "n": 6, "mean": 3.3333333333333335, "axes": { "visualQuality": 3.5, "functionalCompleteness": 3.8333333333333335, "codeQuality": 3.6666666666666665, "polish": 2.3333333333333335 } }, { "journal": "B", "model": "gemma4-e4b-q4", "rubric": "campaign-code-v1", "n": 6, "mean": 3.875, "axes": { "functionalCompleteness": 4, "codeQuality": 3.6666666666666665, "robustness": 3.5, "scopeDiscipline": 4.333333333333333 } }, { "journal": "B", "model": "gemma4-e4b-q4", "rubric": "campaign-data-v1", "n": 2, "mean": 2.125, "axes": { "dataAccuracy": 2, "completeness": 3, "traceability": 1.5, "outputUsability": 2 } }, { "journal": "B", "model": "gemma4-e4b-q4", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.083333333333333, "axes": { "factualAccuracy": 4.083333333333333, "functionalCompleteness": 4.416666666666667, "actionability": 3.5833333333333335, "clarity": 4.25 } }, { "journal": "B", "model": "qwen3.5-9b-q4", "rubric": "default-html-v1", "n": 6, "mean": 2.75, "axes": { "visualQuality": 2.6666666666666665, "functionalCompleteness": 3.3333333333333335, "codeQuality": 3.1666666666666665, "polish": 1.8333333333333333 } }, { "journal": "B", "model": "qwen3.5-9b-q4", "rubric": "campaign-code-v1", "n": 6, "mean": 4.541666666666667, "axes": { "functionalCompleteness": 4.666666666666667, "codeQuality": 5, "robustness": 3.6666666666666665, "scopeDiscipline": 4.833333333333333 } }, { "journal": "B", "model": "qwen3.5-9b-q4", "rubric": "campaign-data-v1", "n": 2, "mean": 3.625, "axes": { "dataAccuracy": 3.5, "completeness": 4.5, "traceability": 3, "outputUsability": 3.5 } }, { "journal": "B", "model": "qwen3.5-9b-q4", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.458333333333333, "axes": { "factualAccuracy": 3.9166666666666665, "functionalCompleteness": 4.75, "actionability": 4.583333333333333, "clarity": 4.583333333333333 } }, { "journal": "B", "model": "ornith1.5-9b-q4", "rubric": "default-html-v1", "n": 6, "mean": 3.75, "axes": { "visualQuality": 4.333333333333333, "functionalCompleteness": 3.6666666666666665, "codeQuality": 3.5, "polish": 3.5 } }, { "journal": "B", "model": "ornith1.5-9b-q4", "rubric": "campaign-code-v1", "n": 5, "mean": 3.75, "axes": { "functionalCompleteness": 3.2, "codeQuality": 4, "robustness": 2.8, "scopeDiscipline": 5 } }, { "journal": "B", "model": "ornith1.5-9b-q4", "rubric": "campaign-data-v1", "n": 1, "mean": 4.75, "axes": { "dataAccuracy": 5, "completeness": 5, "traceability": 5, "outputUsability": 4 } }, { "journal": "B", "model": "ornith1.5-9b-q4", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.458333333333333, "axes": { "factualAccuracy": 4.583333333333333, "functionalCompleteness": 4.166666666666667, "actionability": 4.333333333333333, "clarity": 4.75 } }, { "journal": "B", "model": "qwen3.5-4b-q4", "rubric": "default-html-v1", "n": 6, "mean": 3.3333333333333335, "axes": { "visualQuality": 3.5, "functionalCompleteness": 3.3333333333333335, "codeQuality": 3.6666666666666665, "polish": 2.8333333333333335 } }, { "journal": "B", "model": "qwen3.5-4b-q4", "rubric": "campaign-code-v1", "n": 6, "mean": 4.083333333333333, "axes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 3.6666666666666665, "scopeDiscipline": 4.666666666666667 } }, { "journal": "B", "model": "qwen3.5-4b-q4", "rubric": "campaign-data-v1", "n": 2, "mean": 3, "axes": { "dataAccuracy": 3, "completeness": 3.5, "traceability": 3, "outputUsability": 2.5 } }, { "journal": "B", "model": "qwen3.5-4b-q4", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.125, "axes": { "factualAccuracy": 3.9166666666666665, "functionalCompleteness": 4.166666666666667, "actionability": 4, "clarity": 4.416666666666667 } }, { "journal": "B", "model": "underdog-saluki-27b-iq2-mix", "rubric": "default-html-v1", "n": 6, "mean": 3.4583333333333335, "axes": { "visualQuality": 3.6666666666666665, "functionalCompleteness": 3.5, "codeQuality": 3.6666666666666665, "polish": 3 } }, { "journal": "B", "model": "underdog-saluki-27b-iq2-mix", "rubric": "campaign-code-v1", "n": 6, "mean": 4.666666666666667, "axes": { "functionalCompleteness": 4.666666666666667, "codeQuality": 5.333333333333333, "robustness": 3.6666666666666665, "scopeDiscipline": 5 } }, { "journal": "B", "model": "underdog-saluki-27b-iq2-mix", "rubric": "campaign-data-v1", "n": 2, "mean": 4.5, "axes": { "dataAccuracy": 4.5, "completeness": 4.5, "traceability": 5, "outputUsability": 4 } }, { "journal": "B", "model": "underdog-saluki-27b-iq2-mix", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.854166666666667, "axes": { "factualAccuracy": 4.666666666666667, "functionalCompleteness": 5.083333333333333, "actionability": 4.75, "clarity": 4.916666666666667 } }, { "journal": "B", "model": "qwen3.8-27b-q2", "rubric": "default-html-v1", "n": 6, "mean": 4.125, "axes": { "visualQuality": 4.333333333333333, "functionalCompleteness": 4.166666666666667, "codeQuality": 4.5, "polish": 3.5 } }, { "journal": "B", "model": "qwen3.8-27b-q2", "rubric": "campaign-code-v1", "n": 6, "mean": 5, "axes": { "functionalCompleteness": 5, "codeQuality": 5.666666666666667, "robustness": 4.333333333333333, "scopeDiscipline": 5 } }, { "journal": "B", "model": "qwen3.8-27b-q2", "rubric": "campaign-data-v1", "n": 2, "mean": 4.375, "axes": { "dataAccuracy": 4.5, "completeness": 5, "traceability": 4, "outputUsability": 4 } }, { "journal": "B", "model": "qwen3.8-27b-q2", "rubric": "campaign-doc-v1", "n": 12, "mean": 4.729166666666667, "axes": { "factualAccuracy": 4.5, "functionalCompleteness": 4.916666666666667, "actionability": 4.5, "clarity": 5 } }, { "journal": "N1", "model": "nanbeige4.2-3b-q4", "rubric": "default-html-v1", "n": 2, "mean": 3.625, "axes": { "visualQuality": 4, "functionalCompleteness": 3.5, "codeQuality": 4, "polish": 3 } }, { "journal": "N2", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-code-v1", "n": 1, "mean": 2.75, "axes": { "functionalCompleteness": 2, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 3 } }, { "journal": "N3", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-code-v1", "n": 1, "mean": 4.75, "axes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 } }, { "journal": "N5", "model": "nanbeige4.2-3b-q4", "rubric": "default-html-v1", "n": 2, "mean": 2.875, "axes": { "visualQuality": 3, "functionalCompleteness": 3, "codeQuality": 3.5, "polish": 2 } }, { "journal": "N5", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-doc-v1", "n": 3, "mean": 4.333333333333333, "axes": { "factualAccuracy": 4.666666666666667, "functionalCompleteness": 3.6666666666666665, "actionability": 4.333333333333333, "clarity": 4.666666666666667 } }, { "journal": "N6", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 3.4166666666666665, "axes": { "functionalCompleteness": 3, "codeQuality": 3, "robustness": 3, "scopeDiscipline": 4.666666666666667 } }, { "journal": "N8", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-doc-v1", "n": 2, "mean": 4.625, "axes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4.5, "clarity": 5 } }, { "journal": "N9", "model": "nanbeige4.2-3b-q4", "rubric": "default-html-v1", "n": 1, "mean": 1, "axes": { "visualQuality": 2, "functionalCompleteness": 0, "codeQuality": 1, "polish": 1 } }, { "journal": "N10", "model": "nanbeige4.2-3b-q4", "rubric": "default-html-v1", "n": 1, "mean": 2.5, "axes": { "visualQuality": 3, "functionalCompleteness": 3, "codeQuality": 2, "polish": 2 } }, { "journal": "N10", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-code-v1", "n": 3, "mean": 3.1666666666666665, "axes": { "functionalCompleteness": 3, "codeQuality": 3, "robustness": 2.3333333333333335, "scopeDiscipline": 4.333333333333333 } }, { "journal": "N12", "model": "nanbeige4.2-3b-q4", "rubric": "campaign-doc-v1", "n": 2, "mean": 4.75, "axes": { "factualAccuracy": 5, "functionalCompleteness": 4.5, "actionability": 4.5, "clarity": 5 } } ], "caveats": [ "Unblinded artifact-based inline reviews; distinct rubrics are never averaged into one quality score.", "Missing artifacts receive no invented axes or means. Infrastructure artifact reviews are diagnostic only.", "One original Saluki petshop review exposed 21,664 bytes, exceeding the combined 20 KiB cap by 1,184; its advisory score is excluded. Its deterministic result is unchanged.", "Some early reviews lack contemporaneous content hashes. Current archive hashes cannot prove historical review bytes; verifiedAgainstRecorded retains the distinction." ] }, "performance": { "source": "engine-log", "aggregatesByJournalAndModel": [ { "journal": "A", "model": "gemma4-12b-q4", "n": 7, "pass": 6, "modelFailure": 1, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 7, "fixedEligibleMean": 9.085714285714285, "fixedEligibleN": 7, "nativeTpsMedian": 122.18, "nativeTpsMin": 114.61, "nativeTpsMax": 140.11, "durationMedianMinutes": 1.8593166666666667, "toolCallsMedian": 11 }, { "journal": "A", "model": "gemma4-e4b-q4", "n": 7, "pass": 5, "modelFailure": 2, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 7, "fixedEligibleMean": 8.071428571428571, "fixedEligibleN": 7, "nativeTpsMedian": 204.29, "nativeTpsMin": 176.93, "nativeTpsMax": 218.54, "durationMedianMinutes": 2.0515166666666667, "toolCallsMedian": 8 }, { "journal": "A", "model": "qwen3.5-9b-q4", "n": 7, "pass": 4, "modelFailure": 3, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 7, "fixedEligibleMean": 7.285714285714286, "fixedEligibleN": 7, "nativeTpsMedian": 83.87, "nativeTpsMin": 65.36, "nativeTpsMax": 95.53, "durationMedianMinutes": 2.8505333333333334, "toolCallsMedian": 21 }, { "journal": "A", "model": "ornith1.5-9b-q4", "n": 7, "pass": 5, "modelFailure": 2, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 7, "fixedEligibleMean": 7.871428571428572, "fixedEligibleN": 7, "nativeTpsMedian": 57.2, "nativeTpsMin": 52.7, "nativeTpsMax": 60.94, "durationMedianMinutes": 3.643483333333333, "toolCallsMedian": 14 }, { "journal": "A", "model": "qwen3.5-4b-q4", "n": 7, "pass": 3, "modelFailure": 4, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 7, "fixedEligibleMean": 6.385714285714285, "fixedEligibleN": 7, "nativeTpsMedian": 126.21, "nativeTpsMin": 95.45, "nativeTpsMax": 143.13, "durationMedianMinutes": 2.2227833333333336, "toolCallsMedian": 17 }, { "journal": "A", "model": "underdog-saluki-27b-iq2-mix", "n": 7, "pass": 6, "modelFailure": 1, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 6, "fixedEligibleMean": 9.114285714285714, "fixedEligibleN": 7, "nativeTpsMedian": 12.89, "nativeTpsMin": 12.31, "nativeTpsMax": 13.09, "durationMedianMinutes": 6.437883333333334, "toolCallsMedian": 7 }, { "journal": "A", "model": "qwen3.8-27b-q2", "n": 7, "pass": 5, "modelFailure": 2, "infra": 0, "reviewed": 7, "noArtifact": 0, "advisoryEligible": 7, "fixedEligibleMean": 8.314285714285715, "fixedEligibleN": 7, "nativeTpsMedian": 6.26, "nativeTpsMin": 4.85, "nativeTpsMax": 7.8, "durationMedianMinutes": 10.563216666666667, "toolCallsMedian": 6 }, { "journal": "B", "model": "gemma4-12b-q4", "n": 26, "pass": 21, "modelFailure": 5, "infra": 0, "reviewed": 26, "noArtifact": 2, "advisoryEligible": 24, "fixedEligibleMean": 8.503846153846153, "fixedEligibleN": 26, "nativeTpsMedian": 120.69999999999999, "nativeTpsMin": 11.11, "nativeTpsMax": 155.67, "durationMedianMinutes": 2.0010416666666666, "toolCallsMedian": 8 }, { "journal": "B", "model": "gemma4-e4b-q4", "n": 26, "pass": 20, "modelFailure": 6, "infra": 0, "reviewed": 26, "noArtifact": 0, "advisoryEligible": 26, "fixedEligibleMean": 8.169230769230769, "fixedEligibleN": 26, "nativeTpsMedian": 191.39, "nativeTpsMin": 19.51, "nativeTpsMax": 221.68, "durationMedianMinutes": 1.7631333333333332, "toolCallsMedian": 14.5 }, { "journal": "B", "model": "qwen3.5-9b-q4", "n": 26, "pass": 20, "modelFailure": 6, "infra": 0, "reviewed": 26, "noArtifact": 0, "advisoryEligible": 26, "fixedEligibleMean": 8.492307692307692, "fixedEligibleN": 26, "nativeTpsMedian": 77.25, "nativeTpsMin": 7.5, "nativeTpsMax": 95.94, "durationMedianMinutes": 2.496375, "toolCallsMedian": 16 }, { "journal": "B", "model": "ornith1.5-9b-q4", "n": 26, "pass": 16, "modelFailure": 10, "infra": 0, "reviewed": 26, "noArtifact": 2, "advisoryEligible": 24, "fixedEligibleMean": 7.796153846153847, "fixedEligibleN": 26, "nativeTpsMedian": 56.6, "nativeTpsMin": 4.75, "nativeTpsMax": 64.59, "durationMedianMinutes": 4.987558333333333, "toolCallsMedian": 17.5 }, { "journal": "B", "model": "qwen3.5-4b-q4", "n": 26, "pass": 15, "modelFailure": 11, "infra": 0, "reviewed": 26, "noArtifact": 0, "advisoryEligible": 26, "fixedEligibleMean": 7.380769230769231, "fixedEligibleN": 26, "nativeTpsMedian": 113.07499999999999, "nativeTpsMin": 12.63, "nativeTpsMax": 145.17, "durationMedianMinutes": 1.6998250000000001, "toolCallsMedian": 14.5 }, { "journal": "B", "model": "underdog-saluki-27b-iq2-mix", "n": 26, "pass": 23, "modelFailure": 3, "infra": 0, "reviewed": 26, "noArtifact": 0, "advisoryEligible": 26, "fixedEligibleMean": 9.092307692307692, "fixedEligibleN": 26, "nativeTpsMedian": 12.585, "nativeTpsMin": 2.89, "nativeTpsMax": 13.48, "durationMedianMinutes": 6.375533333333333, "toolCallsMedian": 5 }, { "journal": "B", "model": "qwen3.8-27b-q2", "n": 26, "pass": 21, "modelFailure": 5, "infra": 0, "reviewed": 26, "noArtifact": 0, "advisoryEligible": 26, "fixedEligibleMean": 8.76923076923077, "fixedEligibleN": 26, "nativeTpsMedian": 5.575, "nativeTpsMin": 3.24, "nativeTpsMax": 7.94, "durationMedianMinutes": 15.032733333333333, "toolCallsMedian": 10.5 }, { "journal": "N1", "model": "nanbeige4.2-3b-q4", "n": 3, "pass": 1, "modelFailure": 1, "infra": 1, "reviewed": 3, "noArtifact": 0, "advisoryEligible": 2, "fixedEligibleMean": 5.25, "fixedEligibleN": 2, "nativeTpsMedian": 44.32, "nativeTpsMin": 44.29, "nativeTpsMax": 48.42, "durationMedianMinutes": 24.970016666666666, "toolCallsMedian": 66 }, { "journal": "N2", "model": "nanbeige4.2-3b-q4", "n": 2, "pass": 0, "modelFailure": 1, "infra": 1, "reviewed": 2, "noArtifact": 0, "advisoryEligible": 1, "fixedEligibleMean": 3.1, "fixedEligibleN": 1, "nativeTpsMedian": 51.34, "nativeTpsMin": 43.85, "nativeTpsMax": 58.83, "durationMedianMinutes": 3.2321666666666666, "toolCallsMedian": 44.5 }, { "journal": "N3", "model": "nanbeige4.2-3b-q4", "n": 2, "pass": 1, "modelFailure": 0, "infra": 1, "reviewed": 2, "noArtifact": 1, "advisoryEligible": 1, "fixedEligibleMean": 10, "fixedEligibleN": 1, "nativeTpsMedian": 4.375, "nativeTpsMin": 2.99, "nativeTpsMax": 5.76, "durationMedianMinutes": 35.745200000000004, "toolCallsMedian": 18 }, { "journal": "N4", "model": "nanbeige4.2-3b-q4", "n": 1, "pass": 0, "modelFailure": 0, "infra": 1, "reviewed": 1, "noArtifact": 1, "advisoryEligible": 0, "fixedEligibleMean": null, "fixedEligibleN": 0, "nativeTpsMedian": 2.95, "nativeTpsMin": 2.95, "nativeTpsMax": 2.95, "durationMedianMinutes": 27.0747, "toolCallsMedian": 29 }, { "journal": "N5", "model": "nanbeige4.2-3b-q4", "n": 6, "pass": 3, "modelFailure": 2, "infra": 1, "reviewed": 6, "noArtifact": 1, "advisoryEligible": 5, "fixedEligibleMean": 7.260000000000001, "fixedEligibleN": 5, "nativeTpsMedian": 3.645, "nativeTpsMin": 3.18, "nativeTpsMax": 6.71, "durationMedianMinutes": 44.29505, "toolCallsMedian": 13 }, { "journal": "N6", "model": "nanbeige4.2-3b-q4", "n": 4, "pass": 1, "modelFailure": 2, "infra": 1, "reviewed": 4, "noArtifact": 0, "advisoryEligible": 3, "fixedEligibleMean": 5.866666666666667, "fixedEligibleN": 3, "nativeTpsMedian": 48.67, "nativeTpsMin": 41.91, "nativeTpsMax": 58.05, "durationMedianMinutes": 4.290525, "toolCallsMedian": 54 }, { "journal": "N7", "model": "nanbeige4.2-3b-q4", "n": 1, "pass": 0, "modelFailure": 0, "infra": 1, "reviewed": 1, "noArtifact": 1, "advisoryEligible": 0, "fixedEligibleMean": null, "fixedEligibleN": 0, "nativeTpsMedian": 41.24, "nativeTpsMin": 41.24, "nativeTpsMax": 41.24, "durationMedianMinutes": 7.151633333333334, "toolCallsMedian": 63 }, { "journal": "N8", "model": "nanbeige4.2-3b-q4", "n": 3, "pass": 2, "modelFailure": 0, "infra": 1, "reviewed": 3, "noArtifact": 0, "advisoryEligible": 2, "fixedEligibleMean": 9.75, "fixedEligibleN": 2, "nativeTpsMedian": 43.46, "nativeTpsMin": 41.1, "nativeTpsMax": 45.41, "durationMedianMinutes": 2.2738, "toolCallsMedian": 13 }, { "journal": "N9", "model": "nanbeige4.2-3b-q4", "n": 2, "pass": 0, "modelFailure": 1, "infra": 1, "reviewed": 2, "noArtifact": 0, "advisoryEligible": 1, "fixedEligibleMean": 4.1, "fixedEligibleN": 1, "nativeTpsMedian": 46.379999999999995, "nativeTpsMin": 46.23, "nativeTpsMax": 46.53, "durationMedianMinutes": 19.39480833333333, "toolCallsMedian": 53 }, { "journal": "N10", "model": "nanbeige4.2-3b-q4", "n": 5, "pass": 2, "modelFailure": 2, "infra": 1, "reviewed": 5, "noArtifact": 1, "advisoryEligible": 4, "fixedEligibleMean": 7.05, "fixedEligibleN": 4, "nativeTpsMedian": 44.95, "nativeTpsMin": 40.98, "nativeTpsMax": 55.6, "durationMedianMinutes": 10.122833333333332, "toolCallsMedian": 31 }, { "journal": "N11", "model": "nanbeige4.2-3b-q4", "n": 1, "pass": 0, "modelFailure": 0, "infra": 1, "reviewed": 1, "noArtifact": 1, "advisoryEligible": 0, "fixedEligibleMean": null, "fixedEligibleN": 0, "nativeTpsMedian": 40.45, "nativeTpsMin": 40.45, "nativeTpsMax": 40.45, "durationMedianMinutes": 2.3036333333333334, "toolCallsMedian": 28 }, { "journal": "N12", "model": "nanbeige4.2-3b-q4", "n": 3, "pass": 2, "modelFailure": 0, "infra": 1, "reviewed": 3, "noArtifact": 0, "advisoryEligible": 2, "fixedEligibleMean": 9.75, "fixedEligibleN": 2, "nativeTpsMedian": 42.39, "nativeTpsMin": 41.69, "nativeTpsMax": 43.41, "durationMedianMinutes": 3.8574166666666665, "toolCallsMedian": 19 } ], "caveats": [ "Task performance tables show per-trial engine-log medians, ranges and sample counts; admission-probe fields remain absent.", "System RSS is not total footprint on this discrete GPU. Per-trial GPU memory is reported separately.", "A and B have different runtime hashes. Nanbeige is a separate 32K cohort. Imported cells occur once only.", "N1-N2 ran on Oct 20. N3-N4 observed restricted GPU performance on Oct 22. N5 spans the uncommanded throughput change after 14:00Z. N6-N12 ran after that change. Journals are not a normalized speed comparison.", "A 33 W power cap and 810 MHz memory were observed earlier on Oct 22. Later throughput recovery has no established cause; no operator power/tuning changes were made." ] }, "trials": [ { "cellId": "r1-gemma4-12b-q4-core-tictactoe", "model": "gemma4-12b-q4", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T02:24:29.314Z", "finishedAt": "2026-10-20T02:25:37.860Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 6, "codeQuality": 6, "polish": 4 }, "advisoryMean": 5.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Centered, readable board works at desktop and 390px. Browser replay confirms alternating turns, win/draw detection, repeat-click guards and reset. Clear state/render functions and player colors; fixed sizing, unnamed empty cells, no live status or winning-line highlight keep polish modest.", "reviewInputBytes": 4969, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/5ccab49fa8.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/postmortem.md", "nativeTokensPerSec": 135.24, "nativeGeneratedTokens": 2823, "promptTokensPerSec": 2047.18, "usageSource": "engine-log", "durationMs": 68546, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2975921630859375 }, "peakRssMb": 12806, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-tictactoe", "model": "gemma4-e4b-q4", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T02:27:32.992Z", "finishedAt": "2026-10-20T02:28:27.546Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 6, "codeQuality": 6, "polish": 3 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Win, draw, repeated-cell protection and reset work without page errors. Readable state logic and player colors, but presentation is bare: no visible game title, default restart button and tight status spacing. Empty cells lack accessible names; disableBoard only changes cursors and relies on a separate win guard.", "reviewInputBytes": 4235, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/7847a7a070.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/postmortem.md", "nativeTokensPerSec": 204.91, "nativeGeneratedTokens": 3181, "promptTokensPerSec": 4632.66, "usageSource": "engine-log", "durationMs": 54554, "toolCalls": 5, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2883148193359375 }, "peakRssMb": 6204, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-tictactoe", "model": "qwen3.5-9b-q4", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T02:30:22.564Z", "finishedAt": "2026-10-20T02:33:45.732Z", "success": false, "failureClass": "model", "reason": "repair-aborted: laisa/931ba5bf 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.", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 1, "codeQuality": 3, "polish": 1 }, "advisoryMean": 2, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Every cell click throws clickedCell.getAttribute is not a function: the event listener passes an Event to a handler expecting an element. No marks, turn changes or playable game result. Win/draw/reset code exists but cannot be reached through play. The board is left-aligned apart from the centered heading and restart button.", "reviewInputBytes": 3119, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/970ef8687a.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/postmortem.md", "nativeTokensPerSec": 91.87, "nativeGeneratedTokens": 10322, "promptTokensPerSec": 2523.74, "usageSource": "engine-log", "durationMs": 203168, "toolCalls": 19, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2931060791015625 }, "peakRssMb": 11855, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-tictactoe", "model": "ornith1.5-9b-q4", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T03:31:25.066Z", "finishedAt": "2026-10-20T03:34:34.887Z", "success": true, "failureClass": "pass", "reason": "this-work-e/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent dark styling, colored players, named buttons and a winning highlight. Replay confirms turns, wins, draws and move guards. Reset clears marks but leaves winning cells blue. Fixed 496px board clips at 390px, with scroll width 467px and the left edge offscreen. These state and layout defects reduce otherwise useful polish.", "reviewInputBytes": 4865, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation.", "Browser replay covers one winning line and one draw, not an exhaustive game tree." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/78cefb44ef.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/postmortem.md", "nativeTokensPerSec": 58.88, "nativeGeneratedTokens": 8051, "promptTokensPerSec": 2664.78, "usageSource": "engine-log", "durationMs": 189821, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2889251708984375 }, "peakRssMb": 10142, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-tictactoe", "model": "qwen3.5-4b-q4", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T03:36:21.215Z", "finishedAt": "2026-10-20T03:38:34.582Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 6, "codeQuality": 6, "polish": 4 }, "advisoryMean": 5.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Centered, readable game fits desktop and 390px. Replay confirms alternating turns, wins, draws, occupied-cell/post-win guards and reset without page errors. Straightforward state logic and colored players; fixed-size cells, heavy outlines, unnamed empty buttons and no live status or winning highlight keep it basic.", "reviewInputBytes": 5342, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation.", "Browser replay covers one winning line and one draw, not an exhaustive game tree." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/fe65af7289.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/postmortem.md", "nativeTokensPerSec": 132.66, "nativeGeneratedTokens": 8051, "promptTokensPerSec": 3766.87, "usageSource": "engine-log", "durationMs": 133367, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3100433349609375 }, "peakRssMb": 10225, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-tictactoe", "model": "underdog-saluki-27b-iq2-mix", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T03:40:52.397Z", "finishedAt": "2026-10-20T03:46:00.408Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 6, "codeQuality": 6, "polish": 5 }, "advisoryMean": 5.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clean fluid board fits desktop and 390px, with named buttons, player colors and a green winning line. Win/draw, move guards and reset all work; reset clears the highlight correctly. Small readable functions support the brief. Static cell labels and no live status limit accessibility; presentation remains simple.", "reviewInputBytes": 4407, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation.", "Browser replay covers one winning line and one draw, not an exhaustive game tree." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/a3baf5ab15.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/postmortem.md", "nativeTokensPerSec": 12.92, "nativeGeneratedTokens": 2875, "promptTokensPerSec": 648.41, "usageSource": "engine-log", "durationMs": 308011, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2853851318359375 }, "peakRssMb": 14339, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-tictactoe", "model": "qwen3.8-27b-q2", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T03:48:49.116Z", "finishedAt": "2026-10-20T03:59:22.909Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 7, "functionalCompleteness": 6, "codeQuality": 6, "polish": 6 }, "advisoryMean": 6.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent dark theme, colored markers, winning highlight and styled outcome messages. Replay confirms wins, draws, disabled occupied/finished cells and complete reset. Named buttons and a polite live status add useful accessibility. Readable functions support the brief; the fixed 350px board and static cell labels limit adaptability.", "reviewInputBytes": 4871, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Visuals checked at 960x780 and 390x844; mobile screenshot uses desktop Chromium resized to 390px, not device emulation.", "Browser replay covers one winning line and one draw, not an exhaustive game tree." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/29df2a5d56.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/postmortem.md", "nativeTokensPerSec": 5.98, "nativeGeneratedTokens": 3228, "promptTokensPerSec": 501.87, "usageSource": "engine-log", "durationMs": 633793, "toolCalls": 2, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2719879150390625 }, "peakRssMb": 15339, "peakGpuMemMb": 11600, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-petshop", "model": "gemma4-e4b-q4", "scenario": "petshop", "repetition": 1, "trialId": "petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T04:00:51.727Z", "finishedAt": "2026-10-20T04:03:07.340Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website-build/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 5, "codeQuality": 6, "polish": 3 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "A working local PNG logo and three dynamically rendered pet listings satisfy the basic brief. Desktop and 390px layouts are readable with no page errors. The page is a sparse catalog shell with oversized cards, no contact section or meaningful actions; the generated logo contains garbled lettering. Simple code is easy to follow.", "reviewInputBytes": 2171, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "HTML source read in full within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/9f799cc233.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/postmortem.md", "nativeTokensPerSec": 176.93, "nativeGeneratedTokens": 8776, "promptTokensPerSec": 4317.25, "usageSource": "engine-log", "durationMs": 135613, "toolCalls": 25, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.380096435546875 }, "peakRssMb": 7481, "peakGpuMemMb": 6838, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-petshop", "model": "qwen3.5-9b-q4", "scenario": "petshop", "repetition": 1, "trialId": "petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T04:04:29.764Z", "finishedAt": "2026-10-20T04:10:28.926Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 6, "codeQuality": 6, "polish": 4 }, "advisoryMean": 5.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "A complete static storefront with a valid generated logo, services, priced products, contact details and working section navigation. Layout fits desktop and 390px. Get Directions is a dead # link and product imagery is tiny emoji. The hero depends on an external photo, blocked during this offline review; its online appearance was not assessed.", "reviewInputBytes": 7973, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "HTML source read in full within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/befd0b9f84.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/postmortem.md", "nativeTokensPerSec": 83.87, "nativeGeneratedTokens": 14020, "promptTokensPerSec": 2428.44, "usageSource": "engine-log", "durationMs": 359162, "toolCalls": 37, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.37005615234375 }, "peakRssMb": 15098, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-petshop", "model": "ornith1.5-9b-q4", "scenario": "petshop", "repetition": 1, "trialId": "petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T04:48:11.960Z", "finishedAt": "2026-10-20T04:51:50.569Z", "success": true, "failureClass": "pass", "reason": "build-me-a-single-page-pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 6, "codeQuality": 6, "polish": 4 }, "advisoryMean": 5.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "A coherent orange storefront includes a valid generated logo, services, prices, about/contact content and working section links. Semantic static HTML is straightforward. Desktop spacing is pleasant, but at 390px the unwrapped navigation overlaps the brand and overflows; section content touches screen edges. Placeholder business details remain.", "reviewInputBytes": 5596, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "HTML source read in full within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/33e82f57c1.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/postmortem.md", "nativeTokensPerSec": 59.42, "nativeGeneratedTokens": 5761, "promptTokensPerSec": 2645.3, "usageSource": "engine-log", "durationMs": 218609, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.37274169921875 }, "peakRssMb": 9281, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-petshop", "model": "qwen3.5-4b-q4", "scenario": "petshop", "repetition": 1, "trialId": "petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T04:53:26.551Z", "finishedAt": "2026-10-20T05:02:40.028Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on index.html; missing signals stayed unchanged: working-image, image-asset.", "modelEligible": true, "fixedComposite": 3.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 2, "codeQuality": 3, "polish": 2 }, "advisoryMean": 2.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The PNG logo and all three product images are missing; an unreferenced SVG does not satisfy the image-generation brief. Navigation works, but Shop Now and Add to Cart have no handlers. CSS selectors miss the actual logo/grid classes, leaving oversized stacked cards and default controls. Mobile navigation overflows; the form lacks labels.", "reviewInputBytes": 6302, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/cb442c4b5a.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/postmortem.md", "nativeTokensPerSec": 125.47, "nativeGeneratedTokens": 42265, "promptTokensPerSec": 3699.45, "usageSource": "engine-log", "durationMs": 553477, "toolCalls": 68, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.337249755859375 }, "peakRssMb": 15224, "peakGpuMemMb": 5722, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-petshop", "model": "underdog-saluki-27b-iq2-mix", "scenario": "petshop", "repetition": 1, "trialId": "petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T05:03:48.707Z", "finishedAt": "2026-10-20T05:38:46.246Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judgedWithProtocolDeviation", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 7, "functionalCompleteness": 7, "codeQuality": 6, "polish": 6 }, "advisoryMean": 6.5, "advisoryEligible": false, "advisoryExclusion": "Protocol deviation", "justification": "Consistent warm styling, a generated logo and responsive layout. Category filtering, cart add/remove/totals, mobile menu and contact validation work in replay. Labels and status messages improve usability. Checkout and contact submission only change local state while claiming orders/replies; fictional testimonials and emoji imagery limit production polish.", "reviewInputBytes": 19664, "protocolDeviation": true, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The preliminary source read accidentally exposed all 21664 bytes, exceeding the frozen 20480-byte limit by 1184 bytes. A 19664-byte whitespace-normalized review copy was then saved, but it cannot undo that exposure. This advisory score is disclosed separately and excluded from advisory aggregates; deterministic results are unaffected.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/e957bf320e.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/postmortem.md", "nativeTokensPerSec": 12.6, "nativeGeneratedTokens": 20979, "promptTokensPerSec": 669.43, "usageSource": "engine-log", "durationMs": 2097539, "toolCalls": 26, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.372100830078125 }, "peakRssMb": 18590, "peakGpuMemMb": 10628, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-petshop", "model": "qwen3.8-27b-q2", "scenario": "petshop", "repetition": 1, "trialId": "petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T05:40:18.402Z", "finishedAt": "2026-10-20T06:31:32.554Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"pet-shop-website/workspace/index.html:3:targetnone:frnone:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 5 tool calls (0 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 3.7, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 3, "codeQuality": 6, "polish": 3 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent responsive service page with working navigation and labeled form validation, but both required logo references are broken. The sparse hero and cramped mobile header lack polish. Submission only resets local fields while claiming a message was sent and promising a reply.", "reviewInputBytes": 6385, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/ce09aa4981.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/postmortem.md", "nativeTokensPerSec": 5.46, "nativeGeneratedTokens": 13288, "promptTokensPerSec": 482.18, "usageSource": "engine-log", "durationMs": 3074152, "toolCalls": 20, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.313232421875 }, "peakRssMb": 18530, "peakGpuMemMb": 11606, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-petshop", "model": "gemma4-12b-q4", "scenario": "petshop", "repetition": 1, "trialId": "petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T06:32:43.735Z", "finishedAt": "2026-10-20T06:36:31.690Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 5, "codeQuality": 6, "polish": 4 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clean responsive layout, valid generated paw logo, working section links and labeled form validation. Pet cards are text-only and sparse; the external hero was not assessed online. Contact submission is a local simulation but claims receipt, and feedback lacks a live region.", "reviewInputBytes": 6662, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/4e8ae2df8a.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/postmortem.md", "nativeTokensPerSec": 122.18, "nativeGeneratedTokens": 8438, "promptTokensPerSec": 1856.74, "usageSource": "engine-log", "durationMs": 227955, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.355072021484375 }, "peakRssMb": 20590, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-tankcombat", "model": "qwen3.5-9b-q4", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T06:38:01.838Z", "finishedAt": "2026-10-20T06:39:54.296Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 4, "codeQuality": 3, "polish": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Movement, firing and enemy spawning work; collision/scoring logic exists. But the tank can leave the screen, hits cause no player penalty, and there are no instructions or restart. Tiny rectangles and per-frame random enemy rotation look unfinished. Space release resets the shared spawn/input timer.", "reviewInputBytes": 6480, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/0466a88572.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/postmortem.md", "nativeTokensPerSec": 72.55, "nativeGeneratedTokens": 4862, "promptTokensPerSec": 2378.44, "usageSource": "engine-log", "durationMs": 112458, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.300933837890625 }, "peakRssMb": 9337, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-tankcombat", "model": "ornith1.5-9b-q4", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T08:33:10.916Z", "finishedAt": "2026-10-20T08:46:42.518Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 1, "codeQuality": 3, "polish": 2 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent arcade title/HUD styling, but Space cannot start and no start control is shown. R starts then declares victory within 350ms with zero enemies and score: startWave is never called. Gameplay cannot proceed. Mobile instructions clip, and the unused enemy color picker returns one character.", "reviewInputBytes": 17083, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/c37f5bcf48.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/postmortem.md", "nativeTokensPerSec": 56.37, "nativeGeneratedTokens": 32738, "promptTokensPerSec": 2627.1, "usageSource": "engine-log", "durationMs": 811602, "toolCalls": 11, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2753448486328125 }, "peakRssMb": 9121, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-tankcombat", "model": "qwen3.5-4b-q4", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T08:47:57.752Z", "finishedAt": "2026-10-20T08:49:04.790Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 1, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Enemies animate and shoot, but the player neither moves nor fires. Input is assigned to dataset.inputTick, which becomes the string '[object Object]' instead of a key map; replay confirms unchanged position and zero player bullets. Fixed canvas clips on mobile; no controls are explained.", "reviewInputBytes": 7135, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/6c189d9c1e.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/postmortem.md", "nativeTokensPerSec": 113.17, "nativeGeneratedTokens": 3852, "promptTokensPerSec": 3645.02, "usageSource": "engine-log", "durationMs": 67038, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.284027099609375 }, "peakRssMb": 6711, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-tankcombat", "model": "underdog-saluki-27b-iq2-mix", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T08:50:28.107Z", "finishedAt": "2026-10-20T08:57:48.031Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 2, "codeQuality": 3, "polish": 2 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Arrow movement and enemy activity work with a readable HUD, but Space never fires: fireTimer is NaN and handlers store event.key while firing expects Space. WASD similarly fails. No instructions or usable restart, dead enemies retain depleted HP, and the canvas stretches severely at mobile aspect ratios.", "reviewInputBytes": 8980, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/80664c8d69.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/postmortem.md", "nativeTokensPerSec": 12.94, "nativeGeneratedTokens": 4698, "promptTokensPerSec": 683.15, "usageSource": "engine-log", "durationMs": 439924, "toolCalls": 2, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2756195068359375 }, "peakRssMb": 12899, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-tankcombat", "model": "qwen3.8-27b-q2", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T08:59:10.972Z", "finishedAt": "2026-10-20T09:31:33.790Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"tank-combat-arcade/workspace/index.html:8:targetnone:frnone:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 2 tool calls (0 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 6, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 2, "codeQuality": 3, "polish": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable tank sprites, HUD and controls; movement and firing work initially. The first natural hit throws 'p is not defined', halting the loop with stale HUD and no recovery. Fixed-width canvas clips on mobile. The vocabulary gate failure is separate from this confirmed runtime defect.", "reviewInputBytes": 9947, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/f6fff9016b.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/postmortem.md", "nativeTokensPerSec": 6.61, "nativeGeneratedTokens": 5556, "promptTokensPerSec": 468.4, "usageSource": "engine-log", "durationMs": 1942818, "toolCalls": 5, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2942657470703125 }, "peakRssMb": 17940, "peakGpuMemMb": 11604, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-tankcombat", "model": "gemma4-12b-q4", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T09:32:40.243Z", "finishedAt": "2026-10-20T09:34:02.104Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 0, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The canvas remains entirely blank: draw() is never called and gameLoop() runs only once. Arrow/W input does not move the player; Space adds an unrendered bullet object. Enemy timers cannot produce playable output. The fixed canvas also clips on mobile, despite the official gate pass.", "reviewInputBytes": 6282, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/525d991a7f.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/postmortem.md", "nativeTokensPerSec": 117.45, "nativeGeneratedTokens": 3396, "promptTokensPerSec": 1955.63, "usageSource": "engine-log", "durationMs": 81861, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.33642578125 }, "peakRssMb": 14326, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-tankcombat", "model": "gemma4-e4b-q4", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T09:35:30.485Z", "finishedAt": "2026-10-20T09:38:27.686Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 4, "codeQuality": 3, "polish": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive playfield with working movement and firing, but crude triangle/square graphics and no instructions. Player rotation is invisible, enemies converge on the center, collision only logs Game Over, and there is no health, restart or enemy replenishment. Scoring code was inspected; a scored kill was not replayed.", "reviewInputBytes": 7816, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/22cda23313.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/postmortem.md", "nativeTokensPerSec": 218.54, "nativeGeneratedTokens": 18898, "promptTokensPerSec": 4444.35, "usageSource": "engine-log", "durationMs": 177201, "toolCalls": 14, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.311279296875 }, "peakRssMb": 7299, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-schema-migration", "model": "ornith1.5-9b-q4", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T09:58:55.371Z", "finishedAt": "2026-10-20T10:00:55.094Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable refactor preserves store/handler APIs and correctly splits single- and multi-word names. Migration hard-codes id/email and drops every extra record field, contradicting the brief and its guide. Four tests cover names but miss that loss. Type checking correctly rejects stale User.name access.", "reviewInputBytes": 7301, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/09fc1b7dfc.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/postmortem.md", "nativeTokensPerSec": 60.94, "nativeGeneratedTokens": 4249, "promptTokensPerSec": 2684.59, "usageSource": "engine-log", "durationMs": 119723, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.27978515625 }, "peakRssMb": 9116, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-schema-migration", "model": "qwen3.5-4b-q4", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T10:01:58.016Z", "finishedAt": "2026-10-20T10:03:12.005Z", "success": false, "failureClass": "model", "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 runtime: multi-word: wrong name split (expected {\"firstName\":\"Mary\",\"lastName\":\"Anne Smith\"}, got {\"firstName\":\"Mary\",\"lastName\":\"Anne\"})", "modelEligible": true, "fixedComposite": 4.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 5 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Types, store and handlers are updated while preserving their APIs. Migration loses the remainder of multi-word surnames because split(' ', 2) truncates it, and drops extra fields. Required tests and migration guide are absent. Normal caller typing works and stale User.name access is rejected.", "reviewInputBytes": 2370, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/ced997af06.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/postmortem.md", "nativeTokensPerSec": 143.13, "nativeGeneratedTokens": 1177, "promptTokensPerSec": 3991.95, "usageSource": "engine-log", "durationMs": 73989, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3140716552734375 }, "peakRssMb": 8155, "peakGpuMemMb": 5724, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-schema-migration", "model": "underdog-saluki-27b-iq2-mix", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T10:04:13.361Z", "finishedAt": "2026-10-20T10:10:39.634Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear refactor preserves the store and handler APIs, splits names correctly and supplies the required tests and guide. Migration nevertheless drops every extra record field, contrary to the guide's explicit preservation claim. Type checks and ordinary store operations work; tests miss metadata loss.", "reviewInputBytes": 7050, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/739839695e.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/postmortem.md", "nativeTokensPerSec": 12.77, "nativeGeneratedTokens": 3062, "promptTokensPerSec": 652.93, "usageSource": "engine-log", "durationMs": 386273, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3319244384765625 }, "peakRssMb": 17114, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-schema-migration", "model": "qwen3.8-27b-q2", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T10:11:52.215Z", "finishedAt": "2026-10-20T10:22:01.774Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clean, focused migration preserves extra fields, removes name without mutating input and splits at the first space, including whitespace edge cases. Store and handler APIs and strict source typing work. Three meaningful tests and an accurate guide meet the brief; extra-field preservation is not covered by its own tests.", "reviewInputBytes": 7017, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/4ba2ec696d.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/postmortem.md", "nativeTokensPerSec": 6.99, "nativeGeneratedTokens": 2859, "promptTokensPerSec": 479.29, "usageSource": "engine-log", "durationMs": 609559, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3355865478515625 }, "peakRssMb": 18728, "peakGpuMemMb": 11606, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-schema-migration", "model": "gemma4-12b-q4", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T10:23:10.458Z", "finishedAt": "2026-10-20T10:25:41.427Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Simple refactor preserves APIs and handles name splitting correctly, with three full-record tests and a concise migration guide. The migration copies only id/email and loses all other legacy fields. Source typing and store operations work; stale comments still describe the old name schema.", "reviewInputBytes": 4925, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/56b4f1ce18.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/postmortem.md", "nativeTokensPerSec": 119.57, "nativeGeneratedTokens": 5168, "promptTokensPerSec": 1911.42, "usageSource": "engine-log", "durationMs": 150969, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.390472412109375 }, "peakRssMb": 18034, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-schema-migration", "model": "gemma4-e4b-q4", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T10:26:49.420Z", "finishedAt": "2026-10-20T10:28:52.511Z", "success": false, "failureClass": "model", "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: src/migrate.ts(33,5): error TS2322: Type 'string | undefined' is not assignable to type 'string'.", "modelEligible": true, "fixedComposite": 3.6, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 3, "robustness": 2, "scopeDiscipline": 3 }, "advisoryMean": 2.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Types and display handlers are migrated, but add() returns before storing, so get/list lose new users. Migration fails strict typing, drops extra fields and normalizes whitespace instead of splitting at the first space. Required tests and guide are absent. The premature return introduces an unrelated store regression.", "reviewInputBytes": 2897, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/1cfdb5b3b4.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/postmortem.md", "nativeTokensPerSec": 203.17, "nativeGeneratedTokens": 5828, "promptTokensPerSec": 4443.15, "usageSource": "engine-log", "durationMs": 123091, "toolCalls": 21, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3255767822265625 }, "peakRssMb": 6890, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-schema-migration", "model": "qwen3.5-9b-q4", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T10:29:50.921Z", "finishedAt": "2026-10-20T10:32:41.953Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 3, "scopeDiscipline": 4 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Source handles required name cases and preserves public APIs, but migration drops extra fields. The guide shows a different, broken split(' ', 2) implementation and invents optional input id fields. Sorting also becomes case-insensitive without a requirement. Three tests cover names but miss metadata loss.", "reviewInputBytes": 6202, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/695b7a235c.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/postmortem.md", "nativeTokensPerSec": 80.47, "nativeGeneratedTokens": 4709, "promptTokensPerSec": 2410.02, "usageSource": "engine-log", "durationMs": 171032, "toolCalls": 31, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4970550537109375 }, "peakRssMb": 13330, "peakGpuMemMb": 8104, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-failing-tests-spec", "model": "ornith1.5-9b-q4", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T10:33:45.804Z", "finishedAt": "2026-10-20T10:40:11.351Z", "success": false, "failureClass": "model", "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 2 while the failing detail kept changing; latest checked file src/machine.ts. Last failure: vitest run failed (exit 1): × delivered and cancelled are terminal 4ms | FAIL tests/machine.test.ts > order machine: invalid transitions > delivered and cancelled are terminal | AssertionError: expected [Function] to throw error including 'invalid transition: event \"submit\" is…' but got 'Cannot read properties of undefined (…' || tsc --noEmit failed (exit 2): src/machine.ts(56,29): error TS2345: Argument of type 'string' is not assignable to parameter of type 'Event'.", "modelEligible": true, "fixedComposite": 4.5, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 3, "robustness": 2, "scopeDiscipline": 5 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Most transitions, guards and history work, but delivered is missing from the transition table. Subsequent send/can calls throw an internal undefined.some error instead of the required behavior. Strict compilation also rejects the Set lookup and unchecked find result. Seeded tests/configs remain untouched.", "reviewInputBytes": 2305, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Targeted advisory cases are separate from the official suite and are not exhaustive." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/23265c1535.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/postmortem.md", "nativeTokensPerSec": 55.66, "nativeGeneratedTokens": 10460, "promptTokensPerSec": 2672.87, "usageSource": "engine-log", "durationMs": 385547, "toolCalls": 35, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3280792236328125 }, "peakRssMb": 15426, "peakGpuMemMb": 7256, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-failing-tests-spec", "model": "qwen3.5-4b-q4", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T10:41:10.177Z", "finishedAt": "2026-10-20T10:44:32.364Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on src/machine.ts; missing signals stayed unchanged: vitest-green, tsc-clean. Last failure: vitest run failed (exit 1): × submit moves draft to pending and returns the new state 3ms | × pay moves pending to paid 0ms | × ship moves paid to shipped, deliver moves shipped to delivered 0ms | × cancel is allowed from draft, pending, and paid 1ms | FAIL tests/machine.test.ts > order machine: transitions > submit moves draft to pending and returns the new state | AssertionError: expected 'paid' to be 'pending' // Object.is equality | FAIL tests/machine.test.ts > order machine: transitions > pay moves pending to paid | Error: invalid transition: event \"pay\" is not allowed in state \"paid\" || tsc --noEmit failed (exit 2): src/machine.ts(24,22): error TS2339: Property 'options' does not exist", "modelEligible": true, "fixedComposite": 5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 1, "codeQuality": 2, "robustness": 1, "scopeDiscipline": 3 }, "advisoryMean": 1.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Every accepted non-expedite event goes to paid, so submit, cancel and normal progression are wrong. The options property is absent from the interface and strict compilation fails. Public state/history lose readonly declarations. Seeded tests are intact, but the implementation does not provide a usable lifecycle.", "reviewInputBytes": 1850, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Many state-specific probes cannot reach their setup state because the initial transition is wrong; these are not independent defects." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/3abdb0ce27.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/postmortem.md", "nativeTokensPerSec": 126.21, "nativeGeneratedTokens": 7498, "promptTokensPerSec": 3641.3, "usageSource": "engine-log", "durationMs": 202187, "toolCalls": 29, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3440093994140625 }, "peakRssMb": 13196, "peakGpuMemMb": 5722, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec", "model": "underdog-saluki-27b-iq2-mix", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T10:45:32.020Z", "finishedAt": "2026-10-20T10:47:57.867Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Concise transition table, shared guard logic and intact API meet the ordinary lifecycle tests. However inherited keys such as constructor/toString are accepted as events, replacing state with a function or object instead of rejecting them. Strict typing passes and seeded tests/configs are unchanged.", "reviewInputBytes": 1594, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Targeted advisory cases include unknown object-property names; they do not change the official suite." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/e7bf09d515.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/postmortem.md", "nativeTokensPerSec": 13.09, "nativeGeneratedTokens": 719, "promptTokensPerSec": 688.58, "usageSource": "engine-log", "durationMs": 145847, "toolCalls": 7, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2676239013671875 }, "peakRssMb": 13508, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-failing-tests-spec", "model": "qwen3.8-27b-q2", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T10:48:57.097Z", "finishedAt": "2026-10-20T10:52:32.624Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Well-structured typed transitions and shared guard logic handle ordinary lifecycle, history and rejection cases. Inherited object-property events bypass validation and set state to undefined. Strict typing passes and the seeded specification is untouched, but the public string-event boundary needs an own-property check.", "reviewInputBytes": 1912, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Targeted advisory cases include unknown object-property names; they do not change the official suite." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/e6c6ed3b40.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/postmortem.md", "nativeTokensPerSec": 7.8, "nativeGeneratedTokens": 820, "promptTokensPerSec": 508.69, "usageSource": "engine-log", "durationMs": 215527, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2673492431640625 }, "peakRssMb": 15666, "peakGpuMemMb": 11604, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-failing-tests-spec", "model": "gemma4-12b-q4", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T10:53:31.610Z", "finishedAt": "2026-10-20T10:54:14.416Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable implementation handles ordinary lifecycle and guard cases and returns a copy of the history array. Guard logic is duplicated between can/send. Inherited object-property events are accepted and corrupt state to undefined. Strict typing passes; seeded tests and configs remain unchanged.", "reviewInputBytes": 2185, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Targeted advisory cases include unknown object-property names; they do not change the official suite." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/f52dd22705.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/postmortem.md", "nativeTokensPerSec": 140.11, "nativeGeneratedTokens": 696, "promptTokensPerSec": 2000.01, "usageSource": "engine-log", "durationMs": 42806, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2561187744140625 }, "peakRssMb": 11505, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-failing-tests-spec", "model": "gemma4-e4b-q4", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T10:55:13.257Z", "finishedAt": "2026-10-20T10:55:48.477Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 3, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Ordinary lifecycle and guard cases pass, with the seeded spec unchanged. An unused contradictory transition table, speculative comments and any-typed guards make the code harder to maintain. Inherited object-property events are accepted and corrupt state to undefined. Strict source typing passes.", "reviewInputBytes": 4011, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Targeted advisory cases include unknown object-property names; they do not change the official suite." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/c237b146ec.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/postmortem.md", "nativeTokensPerSec": 204.29, "nativeGeneratedTokens": 1270, "promptTokensPerSec": 4759.98, "usageSource": "engine-log", "durationMs": 35220, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.243621826171875 }, "peakRssMb": 5013, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-failing-tests-spec", "model": "qwen3.5-9b-q4", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T10:56:45.839Z", "finishedAt": "2026-10-20T10:59:18.981Z", "success": false, "failureClass": "model", "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): × expedite is rejected on a machine created without the flag 3ms | FAIL tests/machine.test.ts > order machine: expedite guard > expedite is rejected on a machine created without the flag | AssertionError: expected [Function] to throw an error | - Expected: | + Received:", "modelEligible": true, "fixedComposite": 5.6, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 5 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Most lifecycle and history behavior works, but send ignores the allowed flag that can checks, so a non-expedited order can expedite. Inherited object-property events also corrupt state despite can returning false. Strict typing passes and seeded tests/configs are untouched; duplicated validation diverged.", "reviewInputBytes": 1652, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full changed implementation is below 20 KiB; seeded tests and configs were verified unchanged.", "Reviewed production API was probed across states, events and expedite options, then source was strictly typechecked. No candidate test code was executed.", "Targeted advisory cases include unknown object-property names; they do not change the official suite." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/9e1ae91c78.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/postmortem.md", "nativeTokensPerSec": 95.53, "nativeGeneratedTokens": 4004, "promptTokensPerSec": 2635.94, "usageSource": "engine-log", "durationMs": 153142, "toolCalls": 21, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3296051025390625 }, "peakRssMb": 14072, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-symptom-debug", "model": "qwen3.5-4b-q4", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T11:17:50.509Z", "finishedAt": "2026-10-20T11:19:04.708Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line fix removes the extra cursor increment, preserving all records across varied page sizes and handling tested invalid inputs. Acceptance script and package remain unchanged. The nearby comment still incorrectly says to resume one past the cursor, reducing clarity of an otherwise focused correction.", "reviewInputBytes": 1189, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/a4a2677022.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/postmortem.md", "nativeTokensPerSec": 95.45, "nativeGeneratedTokens": 2875, "promptTokensPerSec": 3511.33, "usageSource": "engine-log", "durationMs": 74199, "toolCalls": 8, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2483673095703125 }, "peakRssMb": 5226, "peakGpuMemMb": 5724, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-symptom-debug", "model": "underdog-saluki-27b-iq2-mix", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T11:20:03.091Z", "finishedAt": "2026-10-20T11:22:25.106Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Minimal correction resumes at the encoded next index and updates the comment to explain that contract. Varied-size walks preserve every record; tested invalid inputs and direct cursors behave correctly. Acceptance script and package remain unchanged, with no unrelated rewrite.", "reviewInputBytes": 1209, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/3a4f70bbe9.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/postmortem.md", "nativeTokensPerSec": 12.89, "nativeGeneratedTokens": 1076, "promptTokensPerSec": 691.39, "usageSource": "engine-log", "durationMs": 142015, "toolCalls": 4, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2327117919921875 }, "peakRssMb": 9853, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-symptom-debug", "model": "qwen3.8-27b-q2", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T11:23:22.606Z", "finishedAt": "2026-10-20T11:27:52.468Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Minimal fix removes the cursor offset and accurately documents that the token already identifies the next unreturned item. Varied-size pagination, invalid inputs and direct cursors pass the targeted checks. Acceptance script and package remain unchanged; scope stays on the defect.", "reviewInputBytes": 1218, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/a4d75ac48f.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/postmortem.md", "nativeTokensPerSec": 4.85, "nativeGeneratedTokens": 947, "promptTokensPerSec": 514.28, "usageSource": "engine-log", "durationMs": 269862, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2337646484375 }, "peakRssMb": 11914, "peakGpuMemMb": 11602, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-symptom-debug", "model": "gemma4-12b-q4", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T11:28:51.094Z", "finishedAt": "2026-10-20T11:30:42.653Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct one-line cursor fix passes varied-size walks, invalid-input checks and direct cursor cases without changing the acceptance script or package. The retained comment contradicts the corrected behavior by saying to resume one past the cursor. Implementation is otherwise focused and preserves existing behavior.", "reviewInputBytes": 1189, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/e3dc1045fd.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/postmortem.md", "nativeTokensPerSec": 114.61, "nativeGeneratedTokens": 758, "promptTokensPerSec": 1679.56, "usageSource": "engine-log", "durationMs": 111559, "toolCalls": 11, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.26031494140625 }, "peakRssMb": 15841, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-symptom-debug", "model": "gemma4-e4b-q4", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T11:31:51.905Z", "finishedAt": "2026-10-20T11:32:26.737Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Removing the extra cursor increment fixes pagination across all targeted walks and boundary checks. Acceptance script and package are untouched. The changed line loses indentation and its preceding comment still describes the old off-by-one behavior; the correction itself remains small and effective.", "reviewInputBytes": 1187, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/b43d6a3154.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/postmortem.md", "nativeTokensPerSec": 200.01, "nativeGeneratedTokens": 2389, "promptTokensPerSec": 4650.34, "usageSource": "engine-log", "durationMs": 34832, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.247833251953125 }, "peakRssMb": 4593, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-symptom-debug", "model": "qwen3.5-9b-q4", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T11:33:23.462Z", "finishedAt": "2026-10-20T11:34:33.091Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The cursor expression now resumes at the correct index, with an equivalent reversed conditional. All targeted pagination, invalid-input and direct-cursor checks pass; acceptance script and package are unchanged. The old comment still falsely describes resuming one position past the cursor.", "reviewInputBytes": 1189, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/f56cd2ab45.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/postmortem.md", "nativeTokensPerSec": 65.36, "nativeGeneratedTokens": 2609, "promptTokensPerSec": 2467.56, "usageSource": "engine-log", "durationMs": 69629, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2476654052734375 }, "peakRssMb": 7479, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-symptom-debug", "model": "ornith1.5-9b-q4", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T11:54:44.142Z", "finishedAt": "2026-10-20T11:57:22.392Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Focused one-line fix removes the cursor offset and updates the comment to explain the next-index contract. Varied-size walks, invalid inputs and direct cursors pass the targeted checks. Acceptance script and package remain unchanged, preserving the scope of the requested bug fix.", "reviewInputBytes": 1226, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final module is below 20 KiB; unchanged fixture code is context, not newly authored work.", "Reviewed production exports passed 62 targeted checks for walks, invalid inputs and direct cursors. No candidate acceptance script was executed.", "Checks are advisory and not exhaustive; source changes remain untouched." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/60812ef699.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/postmortem.md", "nativeTokensPerSec": 57.2, "nativeGeneratedTokens": 7128, "promptTokensPerSec": 2762.87, "usageSource": "engine-log", "durationMs": 158250, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2430877685546875 }, "peakRssMb": 6748, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-data-wrangle", "model": "underdog-saluki-27b-iq2-mix", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T11:58:20.842Z", "finishedAt": "2026-10-20T12:53:36.671Z", "success": false, "failureClass": "model", "reason": "retry loop (long-path): sniff \"data-wrangle:7:targetnone:fr1lbv6pd:rp0:rf0\" stuck for 12m — team idle (2 new tool calls) with no sniff movement for 12m. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 5.3, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 4, "completeness": 5, "traceability": 4, "outputUsability": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 unique customers are present with correct fields and preserved names/ids. Alice's date is wrong (February 11 instead of November 2), and four adjacent email pairs are out of order. Raw inputs are unchanged. A plausible cleaning script exists, but the final JSON disagrees with its date/sort logic.", "reviewInputBytes": 14243, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Independent normalization compared all final rows with the frozen CSV seeds; no candidate scripts were executed.", "The seeded checker and raw data are context, not model-authored artifacts.", "Script behavior was inspected in source only. Final JSON remains the judged deliverable; official gates and performance are unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/539240c020.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/postmortem.md", "nativeTokensPerSec": 12.31, "nativeGeneratedTokens": 36376, "promptTokensPerSec": 548.21, "usageSource": "engine-log", "durationMs": 3315829, "toolCalls": 20, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2336273193359375 }, "peakRssMb": 13323, "peakGpuMemMb": 10628, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-data-wrangle", "model": "qwen3.8-27b-q2", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T12:54:47.061Z", "finishedAt": "2026-10-20T13:54:03.151Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 5, "completeness": 5, "traceability": 4, "outputUsability": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 unique customers match the source-derived newest records, with correct fields, dates and email sorting. Raw inputs are unchanged. The readable final JSON meets the brief. The saved cleaner does not skip the legacy banner and would access undefined emailRaw.trim(), so it cannot reproduce this output as written.", "reviewInputBytes": 14365, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Independent normalization compared every final row with the frozen CSV seeds; candidate scripts were not executed.", "The seeded checker and raw data are context, not model-authored artifacts.", "The cleaner's banner failure is a static source finding: the banner has a nonempty first field and no email field, so it reaches emailRaw.trim().", "Final JSON is the requested deliverable. The helper defect affects advisory traceability only; official gates and performance are unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/6f8acf5e6c.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/postmortem.md", "nativeTokensPerSec": 6.26, "nativeGeneratedTokens": 21443, "promptTokensPerSec": 417.44, "usageSource": "engine-log", "durationMs": 3556090, "toolCalls": 23, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.23468017578125 }, "peakRssMb": 14511, "peakGpuMemMb": 11604, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-data-wrangle", "model": "gemma4-12b-q4", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T13:55:11.837Z", "finishedAt": "2026-10-20T14:02:43.871Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on out/customers.json; missing signals stayed unchanged: row-count. Last failure: expected exactly 56 records after dedup, got 50", "modelEligible": true, "fixedComposite": 5.1, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 4, "traceability": 3, "outputUsability": 3 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The JSON has 50 of 56 customers, eight sort inversions, an incorrect retained Omar id and Robert's older record instead of the newest quoted-name record. Raw inputs are unchanged. Both saved .mjs cleaners use undefined CommonJS require; their intended transform is not reflected in the final data.", "reviewInputBytes": 18257, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Independent normalization compared every final row with the frozen CSV seeds; candidate scripts were not executed.", "The seeded checker and raw data are context, not model-authored artifacts.", "Both .mjs files start with CommonJS require, a static module-format defect. No script was repaired or rerun.", "Final JSON remains the judged deliverable. Official gates, failure classification and performance are unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/776c87db00.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/postmortem.md", "nativeTokensPerSec": 137.95, "nativeGeneratedTokens": 43788, "promptTokensPerSec": 1708.96, "usageSource": "engine-log", "durationMs": 452034, "toolCalls": 29, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 2, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.387451171875 }, "peakRssMb": 18078, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-data-wrangle", "model": "gemma4-e4b-q4", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T14:03:55.034Z", "finishedAt": "2026-10-20T14:06:27.960Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on out/customers.json; missing signals stayed unchanged: unique-emails. Last failure: email \"nils.okafor@example.com\" appears more than once — dedup is missing", "modelEligible": true, "fixedComposite": 4.4, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 3, "traceability": 3, "outputUsability": 2 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The JSON has 41 rows but only 40 unique customers, omitting 16. Nils is duplicated with one wrong id, Una's date is swapped, Sofia's name is altered, and 20 adjacent pairs are unsorted. Raw inputs are unchanged. The selected .mjs cleaner uses undefined require and a parser that truncates unquoted names containing spaces.", "reviewInputBytes": 15765, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Independent normalization compared every final row with the frozen CSV seeds; candidate scripts were not executed.", "The seeded checker and raw data are context, not model-authored artifacts.", "The separate tools/cleanup_data.js was excluded before reading to preserve the frozen input limit. The review does not assess that helper.", "Static findings in scripts/clean_data.mjs: CommonJS require in an ES module, and a CSV field regex that excludes whitespace from unquoted fields.", "Final JSON remains the judged deliverable. Official gates, failure classification and performance are unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/e6f37500aa.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/postmortem.md", "nativeTokensPerSec": 206.21, "nativeGeneratedTokens": 21381, "promptTokensPerSec": 4393.81, "usageSource": "engine-log", "durationMs": 152926, "toolCalls": 8, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3021392822265625 }, "peakRssMb": 5811, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-data-wrangle", "model": "qwen3.5-9b-q4", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T14:07:30.175Z", "finishedAt": "2026-10-20T14:14:26.585Z", "success": false, "failureClass": "model", "reason": "repair-aborted: noor/c997a08b 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.", "modelEligible": true, "fixedComposite": 2.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 4, "completeness": 5, "traceability": 5, "outputUsability": 3 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 expected customers and fields are present with source-derived values, but the 58-row JSON retains both older and newer Omar and Zoe records. Seven adjacent email pairs are unsorted. Names, date conversions and source ids are otherwise faithful; the deliverable still needs deduplication and sorting.", "reviewInputBytes": 15373, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Independent normalization compared every final row with the frozen CSV seeds; no candidate scripts were executed.", "The seeded checker and raw data are review context, not model-authored artifacts. All raw CSVs remained unchanged.", "Write-failure diagnostics include one short rejected-source excerpt and grouped tool outcomes. They explain the failure path without replacing artifact assessment.", "The source-derived older duplicates are reported as mismatches against the required newest records, not invented rows. Official gates and performance are unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/76e4077255.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/postmortem.md", "nativeTokensPerSec": 90.11, "nativeGeneratedTokens": 22051, "promptTokensPerSec": 2291.21, "usageSource": "engine-log", "durationMs": 416410, "toolCalls": 35, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.551910400390625 }, "peakRssMb": 14921, "peakGpuMemMb": 8102, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-data-wrangle", "model": "ornith1.5-9b-q4", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T15:09:41.930Z", "finishedAt": "2026-10-20T16:10:23.242Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on out/customers.json; missing signals stayed unchanged: source-ids-preserved. Last failure: record with id \"L-017\" does not exactly match that source row after the allowed email/date normalization — preserve its source id, name, normalized email, and normalized signupDate together", "modelEligible": true, "fixedComposite": 2.6, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 3, "traceability": 2, "outputUsability": 3 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The valid JSON has 51 of 56 customers, changes Omar and Zoe's source names, and has four sort inversions. Dates and selected source ids otherwise match. Both helper scripts are unusable on the fixture: one misreads the legacy banner and date parts; the other processes headers as records. The final output is not reproducible from them.", "reviewInputBytes": 17616, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "All selected files were read in full within the frozen 20 KiB budget.", "Independent normalization checked every final row against the frozen CSV seeds; all raw CSV files remained unchanged.", "Helper defects are source-inspection findings; candidate scripts were not executed. The seeded checker is context, not model-authored output.", "Official deterministic scores and recorded performance remain unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/8996fa4f7a.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/postmortem.md", "nativeTokensPerSec": 52.7, "nativeGeneratedTokens": 162544, "promptTokensPerSec": 2246.34, "usageSource": "engine-log", "durationMs": 3641312, "toolCalls": 72, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 2, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.319671630859375 }, "peakRssMb": 13826, "peakGpuMemMb": 7256, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-data-wrangle", "model": "qwen3.5-4b-q4", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k", "journal": "A", "journalRoot": "evals/runs/small-model-core-2026-10-19", "runtimeCohort": "Original runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T16:11:32.056Z", "finishedAt": "2026-10-20T16:17:20.947Z", "success": false, "failureClass": "model", "reason": "repair-aborted: noor/ce5e5f2f 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.", "modelEligible": true, "fixedComposite": 2.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 5, "traceability": 3, "outputUsability": 2 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 customers are represented, but 57 rows retain a Robert duplicate and outdated Alice, Omar and Zoe records, with five sort inversions. The helper and rewritten checker use require in .mjs files and cannot run as supplied; the helper also recognizes only a few date patterns. Source values are otherwise preserved.", "reviewInputBytes": 14002, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "All selected files were read in full within the frozen 20 KiB budget.", "Independent normalization checked final rows against the frozen source inputs; all raw CSV files remained unchanged.", "The checker was rewritten by the model and is included as model-authored review evidence; it was not treated as authoritative.", "Candidate scripts were not executed. CommonJS-in-ESM and date-recognition defects were identified from source.", "Official deterministic scores and recorded performance remain unchanged." ], "runDir": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k", "judgmentPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/d7445ade6f.json", "postmortemPath": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/postmortem.md", "nativeTokensPerSec": 130.95, "nativeGeneratedTokens": 30652, "promptTokensPerSec": 3360.1, "usageSource": "engine-log", "durationMs": 348891, "toolCalls": 39, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3606414794921875 }, "peakRssMb": 11393, "peakGpuMemMb": 5724, "runId": "2026-10-20-rtx5070ti-core-original-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-incident-postmortem", "model": "gemma4-12b-q4", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T18:24:36.521Z", "finishedAt": "2026-10-20T18:29:53.429Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core timeout, saturation, peak metrics and recovery are grounded, with all eight sections and cited owners/actions. Some line citations miss their claims; rolling-restart propagation and exact queuing explanations are asserted without evidence. Impact is sparse, the appended deep dive repeats earlier material, and distant action deadlines lack rationale or measurable acceptance criteria.", "reviewInputBytes": 13510, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/bf663c2b52.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/postmortem.md", "nativeTokensPerSec": 108.24, "nativeGeneratedTokens": 20494, "promptTokensPerSec": 1850.89, "usageSource": "engine-log", "durationMs": 316908, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4609375 }, "peakRssMb": 16383, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-incident-postmortem", "model": "gemma4-e4b-q4", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T18:31:27.318Z", "finishedAt": "2026-10-20T18:36:02.668Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 3, "clarity": 2 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Captures the cause and eight sections, but says recovery occurred at14:58, gives inconsistent peak latency and calls14:33 three minutes after14:32. Two near-identical analysis sections and a joined heading expose unfinished repairs. Several actions have TBD deadlines. Claims about missing deployment mechanisms exceed the evidence; repeated core facts do not resolve the contradictions.", "reviewInputBytes": 16332, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/f92e874642.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/postmortem.md", "nativeTokensPerSec": 192.27, "nativeGeneratedTokens": 31520, "promptTokensPerSec": 4051.66, "usageSource": "engine-log", "durationMs": 275350, "toolCalls": 27, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4680023193359375 }, "peakRssMb": 9082, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-incident-postmortem", "model": "qwen3.5-9b-q4", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T18:37:32.319Z", "finishedAt": "2026-10-20T18:40:04.801Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable structure, useful metrics table and specific owned actions. However,15-minute mitigation is wrong, the13.1% peak did not persist until14:53, and errors were3.4%, not below1%, at14:53. It calls45% utilization near capacity and moves the monitoring window before14:58. The report asserts all facts are sourced despite these contradictions and unsupported pipeline claims.", "reviewInputBytes": 15000, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/3ffbbc614f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/postmortem.md", "nativeTokensPerSec": 75.82, "nativeGeneratedTokens": 8059, "promptTokensPerSec": 2373.32, "usageSource": "engine-log", "durationMs": 152482, "toolCalls": 7, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3388519287109375 }, "peakRssMb": 9132, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-incident-postmortem", "model": "ornith1.5-9b-q4", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T18:41:34.271Z", "finishedAt": "2026-10-20T18:48:47.762Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 6, "actionability": 6, "clarity": 6 }, "advisoryMean": 5.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Strong evidence chain, detailed chronology, correct21-minute degradation window, and explicit unknown data/security outcomes. Actions have owners, deadlines and concrete thresholds. It turns absent alert/canary evidence into confident claims that no saturation alert or staged canary existed, and does not clearly reconcile conflicting hotfix-decision timestamps. Some repetition remains.", "reviewInputBytes": 17338, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/379ff8202e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/postmortem.md", "nativeTokensPerSec": 54, "nativeGeneratedTokens": 17708, "promptTokensPerSec": 2527.08, "usageSource": "engine-log", "durationMs": 433491, "toolCalls": 18, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.27703857421875 }, "peakRssMb": 10365, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-incident-postmortem", "model": "qwen3.5-4b-q4", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T18:50:22.998Z", "finishedAt": "2026-10-20T18:53:22.813Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on postmortem.md; missing signals stayed unchanged: no-unsupported-certainty. Last failure: no-unsupported-certainty: unsupported assertion that no data loss occurred", "modelEligible": true, "fixedComposite": 6.2, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 2, "functionalCompleteness": 4, "actionability": 4, "clarity": 3 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core cause and required sections are present, but repeated zero-data-loss/no-security-impact claims are unsupported. Three identical uncertainty notes contradict rather than repair those claims. Response duration, one-minute latency peak and hotfix-start citation are wrong; peer-review bypass and no complaints are also unproven. Actions are usable but the report is not internally reconciled.", "reviewInputBytes": 17074, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4ef8739aa8.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/postmortem.md", "nativeTokensPerSec": 109.5, "nativeGeneratedTokens": 11282, "promptTokensPerSec": 3209.06, "usageSource": "engine-log", "durationMs": 179815, "toolCalls": 16, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4099578857421875 }, "peakRssMb": 8784, "peakGpuMemMb": 5722, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem", "model": "underdog-saluki-27b-iq2-mix", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T18:55:31.591Z", "finishedAt": "2026-10-20T19:06:57.299Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 6, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Explicitly reconciles the hotfix-start source conflict and marks unmeasured impacts unknown. Core cause, recovery and owned actions are useful. Yet it repeatedly calls14:33–14:58 a45-minute interval, reports99% saturation at14:33 when it was97%, and places monitoring before14:58. No-saturation-alert claims are unsupported. These errors weaken an otherwise organized report.", "reviewInputBytes": 16663, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b6e4be8010.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/postmortem.md", "nativeTokensPerSec": 12.19, "nativeGeneratedTokens": 7436, "promptTokensPerSec": 655.07, "usageSource": "engine-log", "durationMs": 685708, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3229217529296875 }, "peakRssMb": 9935, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-incident-postmortem", "model": "qwen3.8-27b-q2", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T19:09:34.032Z", "finishedAt": "2026-10-20T19:57:48.940Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 6, "actionability": 6, "clarity": 5 }, "advisoryMean": 5.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Grounded metrics, correct 21-minute degradation, explicit unknown impacts, and owned actions with dates and thresholds. Reconciles the hotfix source conflict. Its four-minute decision interval is wrong; process/canary claims exceed the sources, and the six-hour merge gap is treated as causal without evidence. Repetition dilutes an otherwise useful report.", "reviewInputBytes": 18861, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full report and four unchanged grounding files were read; only the report is scored as output.", "Deploy.log omitted to respect the combined input cap; claims supported only by that file were not independently checked in this review.", "Direct prose inspection without rendering. Model/source chronology discrepancies were distinguished from arithmetic errors and unsupported claims." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/877ba1515e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/postmortem.md", "nativeTokensPerSec": 5.94, "nativeGeneratedTokens": 16025, "promptTokensPerSec": 462.02, "usageSource": "engine-log", "durationMs": 2894908, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.323974609375 }, "peakRssMb": 15745, "peakGpuMemMb": 11602, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-ops-runbook-anomaly", "model": "gemma4-e4b-q4", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T19:59:56.361Z", "finishedAt": "2026-10-20T20:01:23.490Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on runlog.md; the same scenario failure stayed unchanged. Last failure: runlog.md is missing step 1 (service count)'s observed verification value under its own \"STEP 1\" heading — read state/services.json and count its services array, then record the value verbatim in execution order", "modelEligible": true, "fixedComposite": 5.6, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 3, "actionability": 4, "clarity": 4 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct count, checksum, version and halt values; no later-step files. The runlog omits STEP 4, does not preserve the requested \"14 services\" wording, and duplicates STEP 3. The halt report states both failed conditions but does not clearly record the stopped execution state or provide a complete audit trail.", "reviewInputBytes": 2312, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both produced reports and seven unchanged grounding files were fully read within the combined input cap.", "Sentinel absence was checked on disk. Official trace provenance remains the harness result; no exhaustive session replay.", "Direct prose inspection; no rendered document review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/235da7b09a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/postmortem.md", "nativeTokensPerSec": 206.13, "nativeGeneratedTokens": 3900, "promptTokensPerSec": 4538.52, "usageSource": "engine-log", "durationMs": 87129, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.31280517578125 }, "peakRssMb": 6266, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-ops-runbook-anomaly", "model": "qwen3.5-9b-q4", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T20:02:36.447Z", "finishedAt": "2026-10-20T20:04:14.195Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Complete ordered STEP 1-4 records preserve exact values and source paths. The halt report contrasts fresh/under 24 with stale/71, explains the stop, and leaves steps 5-6 unexecuted; their files are absent. Concise and usable. Minor wording calls the failed check \"Verified\" and cites a vague \"rule 5-6\". Meets the brief without exceptional polish.", "reviewInputBytes": 2787, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both produced reports and seven unchanged grounding files were fully read within the combined input cap.", "Sentinel absence was checked on disk. Official trace provenance remains the harness result; no exhaustive session replay.", "Direct prose inspection; no rendered document review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/cdb69765a3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/postmortem.md", "nativeTokensPerSec": 78.68, "nativeGeneratedTokens": 2607, "promptTokensPerSec": 2506.14, "usageSource": "engine-log", "durationMs": 97748, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.29345703125 }, "peakRssMb": 10773, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-ops-runbook-anomaly", "model": "ornith1.5-9b-q4", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T20:05:25.756Z", "finishedAt": "2026-10-20T20:07:54.976Z", "success": false, "failureClass": "model", "reason": "repair-aborted: onno/1c2de01d 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: runbook.md", "modelEligible": true, "fixedComposite": 5.4, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 6 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct ordered values, STEP 4 failure and stop. The clear table states fresh versus stale and under 24 versus 71; later-step files are absent. The frozen grader rejects the spaced Markdown table; whitespace-only normalization restores its grounding signal. This document assessment does not override the official recovery-aborted outcome.", "reviewInputBytes": 3519, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both produced reports and seven unchanged grounding files were fully read within the combined input cap.", "Sentinel absence was checked. Recorded facts already contain source-reads-grounded; no exhaustive session replay.", "Pure probe compared only halt-report-grounded before/after whitespace normalization, with an empty trace; it is not a replacement full verdict.", "Direct prose inspection; no rendered document review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/77b92dce91.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/postmortem.md", "nativeTokensPerSec": 58.57, "nativeGeneratedTokens": 3652, "promptTokensPerSec": 2672.5, "usageSource": "engine-log", "durationMs": 149220, "toolCalls": 31, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.333770751953125 }, "peakRssMb": 11469, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-ops-runbook-anomaly", "model": "qwen3.5-4b-q4", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T20:09:11.439Z", "finishedAt": "2026-10-20T20:10:49.484Z", "success": false, "failureClass": "model", "reason": "repair-aborted: onno/2252265f 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: runbook.md", "modelEligible": true, "fixedComposite": 3.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All four steps contain correct observed values in order, including stale/71 against fresh/under 24. The halt report identifies the failed check and source; later-step files are absent. Plain STEP labels lack Markdown heading markers, so the frozen parser misses the records. The halt report does not explicitly say steps 5-6 were skipped. Useful content with incomplete presentation.", "reviewInputBytes": 2835, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both produced reports and seven unchanged grounding files read in full. Only reports are scored output.", "Sentinel absence checked. Frozen parser requires Markdown STEP headings; plain labels are not recognized. Official score and failure class unchanged.", "Direct prose inspection; no rendered-document or exhaustive session replay." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/57d232eb86.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/postmortem.md", "nativeTokensPerSec": 115.65, "nativeGeneratedTokens": 3465, "promptTokensPerSec": 3749.03, "usageSource": "engine-log", "durationMs": 98045, "toolCalls": 31, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3125762939453125 }, "peakRssMb": 10075, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly", "model": "underdog-saluki-27b-iq2-mix", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T20:12:04.942Z", "finishedAt": "2026-10-20T20:15:35.401Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact STEP 1-4 values appear in order. The halt report gives the source, failing step, both expected conditions and observed stale/71 values, and explicitly leaves steps 5-6 unexecuted. Their files are absent. A minimal, correct audit record that meets the brief.", "reviewInputBytes": 2169, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both generated reports and all seven unchanged seed files were read. Sentinel absence checked on disk.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2b8e169b31.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/postmortem.md", "nativeTokensPerSec": 12.74, "nativeGeneratedTokens": 1398, "promptTokensPerSec": 616.8, "usageSource": "engine-log", "durationMs": 210459, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2583465576171875 }, "peakRssMb": 14628, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-ops-runbook-anomaly", "model": "qwen3.8-27b-q2", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T20:16:42.516Z", "finishedAt": "2026-10-20T20:23:29.431Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 6 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct ordered verification values and a clear halt report. It explains both failed checks, names the omitted downstream files, and summarizes the successfully completed steps with sources. No later-step outputs exist. Readable and easy to audit; some status information is repeated across the two documents.", "reviewInputBytes": 2654, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both generated reports and all seven unchanged seed files were read. Sentinel absence checked on disk.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/ee03a8d6c7.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/postmortem.md", "nativeTokensPerSec": 5.5, "nativeGeneratedTokens": 1556, "promptTokensPerSec": 450.09, "usageSource": "engine-log", "durationMs": 406915, "toolCalls": 12, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2584381103515625 }, "peakRssMb": 16594, "peakGpuMemMb": 11602, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-ops-runbook-anomaly", "model": "gemma4-12b-q4", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T20:24:44.707Z", "finishedAt": "2026-10-20T20:26:30.632Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "A short, complete ordered runlog and grounded halt report preserve the exact values. Expected fresh/under24 and observed stale/71 are clearly separated; steps 5-6 are explicitly not executed and their files are absent. The minimal presentation meets the brief.", "reviewInputBytes": 2074, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Both generated reports and all seven unchanged seed files were read. Sentinel absence checked on disk.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/32c509ac38.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/postmortem.md", "nativeTokensPerSec": 130.93, "nativeGeneratedTokens": 1353, "promptTokensPerSec": 1860.82, "usageSource": "engine-log", "durationMs": 105925, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.317718505859375 }, "peakRssMb": 17795, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-plan-and-estimate", "model": "qwen3.5-9b-q4", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T20:27:47.935Z", "finishedAt": "2026-10-20T20:28:40.833Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct roster owners, floor-plan sign-off before booking, and clear done-states. Packing and workshop work are planned, but there is no actual transport/unpack task; final readiness does not depend on packing. A vendor same-day-expedite SLA is asserted without support. The useful structure needs a more complete execution chain.", "reviewInputBytes": 4411, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7a48e4d339.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/postmortem.md", "nativeTokensPerSec": 52.78, "nativeGeneratedTokens": 1119, "promptTokensPerSec": 2608.23, "usageSource": "engine-log", "durationMs": 52898, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2496185302734375 }, "peakRssMb": 7364, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-plan-and-estimate", "model": "ornith1.5-9b-q4", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T20:29:48.644Z", "finishedAt": "2026-10-20T20:30:53.418Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 3, "actionability": 3, "clarity": 5 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Preserves the booking constraint and roster, with readable tasks and network checks. It explicitly leaves the workshop build without an owner, although Sanne is available. Packing appears only in a completion condition; no physical move execution is planned. Eight rows satisfy the structural gate while material work remains unassigned.", "reviewInputBytes": 4183, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/3d6f1c04b4.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/postmortem.md", "nativeTokensPerSec": 59.4, "nativeGeneratedTokens": 2071, "promptTokensPerSec": 2766.62, "usageSource": "engine-log", "durationMs": 64774, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.24505615234375 }, "peakRssMb": 6767, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-plan-and-estimate", "model": "qwen3.5-4b-q4", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T20:32:02.155Z", "finishedAt": "2026-10-20T20:32:49.795Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 3, "clarity": 4 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Detailed roster-owned tasks, but T1 receives a signed building plan before seating/workshop design; booking depends on that instead of approval of the completed layout. External internet activation and workshop execution are not assigned. Print/scan testing is outside the final verification dependency chain. Extra detail does not close these gaps.", "reviewInputBytes": 5577, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2fd3230e10.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/postmortem.md", "nativeTokensPerSec": 79.54, "nativeGeneratedTokens": 1613, "promptTokensPerSec": 3776.37, "usageSource": "engine-log", "durationMs": 47640, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2503204345703125 }, "peakRssMb": 4949, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate", "model": "underdog-saluki-27b-iq2-mix", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T20:33:56.625Z", "finishedAt": "2026-10-20T20:36:19.340Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Covers the signed layout, booking, physical move, workshop build with an internal owner, and tested network/print setup. Readiness depends on the completed workstreams; assumptions are explicit and done-states observable. A concise usable plan. The proposed network-before-move ordering exists only in risk prose, and timing buffers remain broad.", "reviewInputBytes": 3823, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/101e942882.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/postmortem.md", "nativeTokensPerSec": 13.14, "nativeGeneratedTokens": 1162, "promptTokensPerSec": 694.82, "usageSource": "engine-log", "durationMs": 142715, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.23468017578125 }, "peakRssMb": 9723, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-plan-and-estimate", "model": "qwen3.8-27b-q2", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T20:37:26.850Z", "finishedAt": "2026-10-20T20:44:44.571Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Useful logistics, owned tasks and concrete tests. Sign-off can be satisfied by returned changes rather than clear approval; final readiness does not depend on T7 verification. Workshop timing conflicts between table and risk text, while a vendor SLA is assumed. Moving after network setup is sound, but the dependency chain needs tightening.", "reviewInputBytes": 5794, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2d64808227.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/postmortem.md", "nativeTokensPerSec": 3.94, "nativeGeneratedTokens": 1450, "promptTokensPerSec": 498.97, "usageSource": "engine-log", "durationMs": 437721, "toolCalls": 4, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2357330322265625 }, "peakRssMb": 11770, "peakGpuMemMb": 11602, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-plan-and-estimate", "model": "gemma4-12b-q4", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T20:45:43.126Z", "finishedAt": "2026-10-20T20:46:25.326Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear roster ownership, signed-layout booking dependency, workshop execution and physical move. Network testing is stated before move day. Print/scan hardware is ordered without an explicit installation/test task, and no final readiness gate checks all workstreams. Mitigations are mostly general reminders.", "reviewInputBytes": 3300, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d5dcb23be8.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/postmortem.md", "nativeTokensPerSec": 103.47, "nativeGeneratedTokens": 1428, "promptTokensPerSec": 1885.53, "usageSource": "engine-log", "durationMs": 42200, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.262298583984375 }, "peakRssMb": 10677, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-plan-and-estimate", "model": "gemma4-e4b-q4", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T20:47:24.044Z", "finishedAt": "2026-10-20T20:48:48.504Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on plan.md; the same scenario failure stayed unchanged. Last failure: plan table: no Work plan row books the movers", "modelEligible": true, "fixedComposite": 6.4, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 3, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The moving company is explicitly booked after sign-off; a synonym-only probe confirms the frozen grader misses that wording. Internal owners and workshop execution are present. IT/design roles are poorly matched, print/scan has no dependencies and is omitted from final readiness, and risks lack concrete mitigations. Official failure is preserved.", "reviewInputBytes": 4013, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/06a1853983.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/postmortem.md", "nativeTokensPerSec": 205.42, "nativeGeneratedTokens": 7714, "promptTokensPerSec": 4512.06, "usageSource": "engine-log", "durationMs": 84460, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.279388427734375 }, "peakRssMb": 6318, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-ornith1.5-9b-q4-core-conflict-synthesis", "model": "ornith1.5-9b-q4", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T20:49:46.564Z", "finishedAt": "2026-10-20T20:55:24.450Z", "success": false, "failureClass": "model", "reason": "repair-aborted: tamsin/9178cab3 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.", "modelEligible": true, "fixedComposite": 5.3, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 3, "actionability": 4, "clarity": 4 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core date, budget, ownership and source priority are correct, with explicit conflict resolutions and useful questions. However, the superseded240,000 budget escapes into Open questions despite a claim that all old values are quarantined. Repeated budget items and self-descriptions add noise. The artifact does not meet the reconciliation boundary.", "reviewInputBytes": 4494, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/422b1d0552.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/postmortem.md", "nativeTokensPerSec": 56.73, "nativeGeneratedTokens": 13114, "promptTokensPerSec": 2609.65, "usageSource": "engine-log", "durationMs": 337886, "toolCalls": 47, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4822998046875 }, "peakRssMb": 13961, "peakGpuMemMb": 7256, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-4b-q4-core-conflict-synthesis", "model": "qwen3.5-4b-q4", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T20:56:24.261Z", "finishedAt": "2026-10-20T20:58:16.775Z", "success": false, "failureClass": "model", "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", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winners and three clear conflict resolutions; old values stay in the conflicts table. The plan is concise and its open questions are grounded. Most plan claims use bare or informal source names, leaving only three claims cited in the requested backticked form. Budget line-item allocation is omitted. Useful, but incomplete citation compliance.", "reviewInputBytes": 2779, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete generated document and all unchanged grounding sources were read; only generated output is scored.", "Direct prose, facts and dependency review; no rendered-document inspection or exhaustive session replay.", "Frozen gate, classification and score remain unchanged; any pure-evaluator probes are diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/e092513fee.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/postmortem.md", "nativeTokensPerSec": 121.52, "nativeGeneratedTokens": 5728, "promptTokensPerSec": 3814.98, "usageSource": "engine-log", "durationMs": 112514, "toolCalls": 36, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3670196533203125 }, "peakRssMb": 8273, "peakGpuMemMb": 5724, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis", "model": "underdog-saluki-27b-iq2-mix", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T20:59:19.545Z", "finishedAt": "2026-10-20T21:22:24.565Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The three required conflicts are resolved with sources and old values stay quarantined. Budget allocation and ownership are clear. However, it invents a contractor-count conflict: a spend line does not contradict two contractors. It also assumes the six-week migration ends on launch day. Extra reconciliation prose and citation machinery obscure a mostly useful synthesis.", "reviewInputBytes": 4884, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete output and six unchanged sources read directly. No rendered-document inspection or exhaustive session replay.", "Fixed gate, score and classification remain unchanged. Any existing-checker probe is diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d2f7712840.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/postmortem.md", "nativeTokensPerSec": 12.31, "nativeGeneratedTokens": 14375, "promptTokensPerSec": 661.88, "usageSource": "engine-log", "durationMs": 1385020, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2771759033203125 }, "peakRssMb": 18089, "peakGpuMemMb": 10628, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.8-27b-q2-core-conflict-synthesis", "model": "qwen3.8-27b-q2", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T21:23:33.476Z", "finishedAt": "2026-10-20T21:44:11.695Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct source priorities, values, budget breakdown and quarantined conflicts with frequent citations. It carries Tuesday meetings into the plan while calling their validity unverified, and says no source confirms contractor count despite the product memo naming two. The useful structure needs a cleaner separation of provisional assumptions from the adopted plan.", "reviewInputBytes": 4034, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete output and six unchanged sources read directly. No rendered-document inspection or exhaustive session replay.", "Fixed gate, score and classification remain unchanged. Any existing-checker probe is diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b273bb308f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/postmortem.md", "nativeTokensPerSec": 5.65, "nativeGeneratedTokens": 6261, "promptTokensPerSec": 478.59, "usageSource": "engine-log", "durationMs": 1238219, "toolCalls": 15, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.285003662109375 }, "peakRssMb": 16789, "peakGpuMemMb": 11604, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-12b-q4-core-conflict-synthesis", "model": "gemma4-12b-q4", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T21:45:20.755Z", "finishedAt": "2026-10-20T21:46:27.055Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 4, "clarity": 5 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Concise core plan, correct winners, budget breakdown and source-linked resolutions. Open Questions repeats the superseded budget as240k outside the conflicts section and assumes a budget reduction. The frozen matcher misses the shorthand; expanding it in memory exposes the quarantine failure. Official PASS is preserved, but the artifact violates the explicit boundary.", "reviewInputBytes": 2466, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete output and six unchanged sources read directly. No rendered-document inspection or exhaustive session replay.", "Fixed gate, score and classification remain unchanged. Any existing-checker probe is diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/96d61bd46c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/postmortem.md", "nativeTokensPerSec": 129.97, "nativeGeneratedTokens": 1795, "promptTokensPerSec": 1821.28, "usageSource": "engine-log", "durationMs": 66300, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.353424072265625 }, "peakRssMb": 14003, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-gemma4-e4b-q4-core-conflict-synthesis", "model": "gemma4-e4b-q4", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T21:47:52.309Z", "finishedAt": "2026-10-20T21:49:45.720Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 4, "actionability": 3, "clarity": 5 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winning date, total budget, owner and three source-priority resolutions. The plan changes EUR line items to dollar signs, omits the contractor amount there, and dismisses historical weekly syncs as irrelevant without evidence. Numbered citations have no reference list. Clear structure, but several details need correction or explicit uncertainty.", "reviewInputBytes": 3083, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete output and six unchanged sources read directly. No rendered-document inspection or exhaustive session replay.", "Fixed gate, score and classification remain unchanged. Any existing-checker probe is diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/dd5866a97d.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/postmortem.md", "nativeTokensPerSec": 210.3, "nativeGeneratedTokens": 9778, "promptTokensPerSec": 4712, "usageSource": "engine-log", "durationMs": 113411, "toolCalls": 19, "redFlags": [ { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/sessions/tamsin", "tool": "(no write_file this turn)", "argsSummary": "content_chars=3023 h2_count=3 preview=\"```markdown # Skylark Launch Synthesis ## Summary This brief synthesizes information from the product, engineering, fin…\"", "atTurn": 1, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." } ], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.31451416015625 }, "peakRssMb": 7028, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-qwen3.5-9b-q4-core-conflict-synthesis", "model": "qwen3.5-9b-q4", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T21:50:55.683Z", "finishedAt": "2026-10-20T21:52:17.943Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear consolidated values, complete budget allocation and correctly quarantined conflict resolutions. Sources support the core claims. It invents a need to reassign weekly syncs specifically to Marcus and Iris rather than ask whether the old cadence remains valid. Extra non-conflicts and duplicated Sources sections reduce concision.", "reviewInputBytes": 3377, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete output and six unchanged sources read directly. No rendered-document inspection or exhaustive session replay.", "Fixed gate, score and classification remain unchanged. Any existing-checker probe is diagnostic only." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6d35f8b382.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/postmortem.md", "nativeTokensPerSec": 81.55, "nativeGeneratedTokens": 3174, "promptTokensPerSec": 2458.78, "usageSource": "engine-log", "durationMs": 82260, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3355865478515625 }, "peakRssMb": 8416, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-tictactoe", "model": "gemma4-e4b-q4", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-20T21:53:25.608Z", "finishedAt": "2026-10-20T21:54:27.560Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 3 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable large board and status, with correct alternating turns, repeat-click protection, row win, post-win lock, draw and keyboard activation. No page errors or narrow-screen overflow. The requested basic game works, but another round requires reloading; there is no reset, win highlight or accessible cell names. Plain styling and spare feedback limit polish.", "reviewInputBytes": 3728, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Final artifact selected from facts.sniff.latest.filePath, not first HTML found. Full selected source read within20KiB.", "Offline desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation.", "Targeted row-win, draw, repeat-click, terminal-lock, keyboard and reset checks are not exhaustive. When no reset exists, page reload was used only to start the next probe.", "Desktop and narrow screenshots visually inspected; Ornith winning-state screenshot also inspected. No external judge API." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/109deb2e74.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/postmortem.md", "nativeTokensPerSec": 204.25, "nativeGeneratedTokens": 3376, "promptTokensPerSec": 4738.97, "usageSource": "engine-log", "durationMs": 61952, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2878265380859375 }, "peakRssMb": 6168, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-tictactoe", "model": "qwen3.5-9b-q4", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T21:55:56.780Z", "finishedAt": "2026-10-20T21:59:43.445Z", "success": true, "failureClass": "pass", "reason": "build-tic-tac-toe-game/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))", "modelEligible": true, "fixedComposite": 7.8, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 5, "codeQuality": 4, "polish": 2 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The declared project artifact plays correctly in row-win, draw, repeat-click and post-win checks, with keyboard buttons and no errors. It is a small default-styled board with no reset, marked-cell labels or winner highlighting. Viewport metadata is absent. Fits a narrow desktop viewport, but actual mobile scaling was not tested.", "reviewInputBytes": 2339, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Final artifact selected from facts.sniff.latest.filePath, not first HTML found. Full selected source read within20KiB.", "Offline desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation.", "Targeted row-win, draw, repeat-click, terminal-lock, keyboard and reset checks are not exhaustive. When no reset exists, page reload was used only to start the next probe.", "Desktop and narrow screenshots visually inspected; Ornith winning-state screenshot also inspected. No external judge API." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d1ac09efb4.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/postmortem.md", "nativeTokensPerSec": 85.71, "nativeGeneratedTokens": 11071, "promptTokensPerSec": 2483.25, "usageSource": "engine-log", "durationMs": 226665, "toolCalls": 22, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.299346923828125 }, "peakRssMb": 12607, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-tictactoe", "model": "ornith1.5-9b-q4", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T22:01:31.805Z", "finishedAt": "2026-10-20T22:04:18.416Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 5, "codeQuality": 5, "polish": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent centered dark layout, colored marks and a highlighted winning line. Turn changes, duplicate-click guard, terminal lock, draw and visible New Game reset all work without errors; reset clears highlights. Fits both inspected widths and supports keyboard buttons. Blue status text is subdued against the dark background; cell labels do not announce changing marks.", "reviewInputBytes": 4900, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Final artifact selected from facts.sniff.latest.filePath, not first HTML found. Full selected source read within20KiB.", "Offline desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation.", "Targeted row-win, draw, repeat-click, terminal-lock, keyboard and reset checks are not exhaustive. When no reset exists, page reload was used only to start the next probe.", "Desktop and narrow screenshots visually inspected; Ornith winning-state screenshot also inspected. No external judge API." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/616e16a003.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/postmortem.md", "nativeTokensPerSec": 59.11, "nativeGeneratedTokens": 7618, "promptTokensPerSec": 2768.61, "usageSource": "engine-log", "durationMs": 166611, "toolCalls": 4, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2783355712890625 }, "peakRssMb": 8520, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-tictactoe", "model": "qwen3.5-4b-q4", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T22:05:52.663Z", "finishedAt": "2026-10-20T22:06:49.134Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clean centered board with distinct mark colors and a working Restart Game control. Alternation, repeat-click protection, row win, post-win lock, draw and reset all work without errors or narrow-screen overflow. Straightforward code and layout meet the brief. No winning-line emphasis, changing cell labels or live status announcement.", "reviewInputBytes": 4713, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Final artifact selected from facts.sniff.latest.filePath, not first HTML found. Full selected source read within20KiB.", "Offline desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation.", "Targeted row-win, draw, repeat-click, terminal-lock, keyboard and reset checks are not exhaustive. When no reset exists, page reload was used only to start the next probe.", "Desktop and narrow screenshots visually inspected; Ornith winning-state screenshot also inspected. No external judge API." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d197c56107.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/postmortem.md", "nativeTokensPerSec": 125.49, "nativeGeneratedTokens": 2005, "promptTokensPerSec": 3992.39, "usageSource": "engine-log", "durationMs": 56471, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2839508056640625 }, "peakRssMb": 6901, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-tictactoe", "model": "underdog-saluki-27b-iq2-mix", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T22:08:23.183Z", "finishedAt": "2026-10-20T22:12:18.282Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 5, "codeQuality": 5, "polish": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Consistent dark palette, clear status badge, colored marks and winning-line highlight. Alternation, duplicate-click guard, row win, post-win lock, draw, keyboard activation and New Game reset work without errors. Fits390px; excess body height creates needless vertical scroll. Cell names stay generic and status changes are not announced. A complete basic game.", "reviewInputBytes": 4131, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected source within20KiB. Artifact selected from facts.sniff.latest.filePath.", "Offline desktop Chromium1200x900/390x844, not mobile-device emulation.", "Targeted row-win/draw/reset/keyboard probes are not an exhaustive game-tree test. No official result or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/89e452db6a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/postmortem.md", "nativeTokensPerSec": 13.23, "nativeGeneratedTokens": 2091, "promptTokensPerSec": 689.51, "usageSource": "engine-log", "durationMs": 235099, "toolCalls": 2, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.27093505859375 }, "peakRssMb": 12286, "peakGpuMemMb": 10624, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-tictactoe", "model": "qwen3.8-27b-q2", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T22:13:51.670Z", "finishedAt": "2026-10-20T22:20:59.785Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 5, "codeQuality": 5, "polish": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear board, distinct X/O colors and visible winning cells. Turns, repeat-click protection, win lock, draw, keyboard activation and reset pass targeted replay at desktop and narrow widths. Simple readable state logic meets the brief; blank cells lack accessible names and status changes are not announced.", "reviewInputBytes": 3356, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8d0d800bcd.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/postmortem.md", "nativeTokensPerSec": 6.3, "nativeGeneratedTokens": 2058, "promptTokensPerSec": 476.33, "usageSource": "engine-log", "durationMs": 428115, "toolCalls": 2, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2719879150390625 }, "peakRssMb": 14408, "peakGpuMemMb": 11600, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-tictactoe", "model": "gemma4-12b-q4", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-20T22:22:48.785Z", "finishedAt": "2026-10-20T22:23:45.758Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Basic centered board fits both widths and passes turn, repeated-click, win lock, draw, reset and keyboard probes without errors. State handling is straightforward. Plain monochrome marks and a default-style reset button offer little finish; winning cells are not highlighted, empty buttons lack labels and status is not live.", "reviewInputBytes": 4567, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d74277a4e7.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/postmortem.md", "nativeTokensPerSec": 155.67, "nativeGeneratedTokens": 1701, "promptTokensPerSec": 1972.22, "usageSource": "engine-log", "durationMs": 56973, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2975921630859375 }, "peakRssMb": 12025, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-petshop", "model": "qwen3.5-9b-q4", "scenario": "petshop", "repetition": 2, "trialId": "petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-20T22:25:20.906Z", "finishedAt": "2026-10-20T22:33:07.828Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 4, "codeQuality": 4, "polish": 2 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Pet-shop sections and local PNG load, but the header uses a clipped SVG while the generated logo appears as the Dogs product image with garbled lettering. View buttons only alert. External placeholder images fail in offline review; live availability is untested. Layout narrows without overflow, but viewport metadata is missing and branding is unfinished.", "reviewInputBytes": 3492, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d9806fb77b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/postmortem.md", "nativeTokensPerSec": 67.44, "nativeGeneratedTokens": 15954, "promptTokensPerSec": 2334.56, "usageSource": "engine-log", "durationMs": 466922, "toolCalls": 77, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5029754638671875 }, "peakRssMb": 17257, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-petshop", "model": "ornith1.5-9b-q4", "scenario": "petshop", "repetition": 2, "trialId": "petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 3, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-20T22:35:00.987Z", "finishedAt": "2026-10-20T22:40:28.325Z", "success": true, "failureClass": "pass", "reason": "build-me-a-single-page-pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 6, "functionalCompleteness": 5, "codeQuality": 5, "polish": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent cream/orange layout, readable service/product cards, local logo and contact section meet the landing-page brief. Anchors resolve and mobile menu opens, navigates and closes with updated aria-expanded. Structured CSS/HTML are readable. Small detailed logo and low-contrast white/orange CTA text limit finish; hidden menu links remain focusable.", "reviewInputBytes": 10672, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5db999fd9b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/postmortem.md", "nativeTokensPerSec": 56.59, "nativeGeneratedTokens": 10630, "promptTokensPerSec": 2658.6, "usageSource": "engine-log", "durationMs": 327338, "toolCalls": 21, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.35809326171875 }, "peakRssMb": 9927, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-petshop", "model": "qwen3.5-4b-q4", "scenario": "petshop", "repetition": 2, "trialId": "petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-20T22:42:13.847Z", "finishedAt": "2026-10-20T22:48:40.884Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on index.html; missing signals stayed unchanged: working-image, image-asset.", "modelEligible": true, "fixedComposite": 3.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 3, "codeQuality": 4, "polish": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Six priced products and responsive cards are usable; add/remove totals, empty-checkout handling and the explicitly demo checkout work. Required PNG is absent, so the header displays a broken image despite an unused SVG. Initial empty cart has no message, toast updates lack live announcements, and generic emoji cards leave the site unfinished.", "reviewInputBytes": 10495, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4803b2395c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/postmortem.md", "nativeTokensPerSec": 120.76, "nativeGeneratedTokens": 24981, "promptTokensPerSec": 3632.27, "usageSource": "engine-log", "durationMs": 387037, "toolCalls": 56, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.59478759765625 }, "peakRssMb": 16177, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-petshop", "model": "underdog-saluki-27b-iq2-mix", "scenario": "petshop", "repetition": 2, "trialId": "petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-20T22:49:49.529Z", "finishedAt": "2026-10-20T23:11:12.447Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Local generated logo, coherent catalog and responsive layout meet the brief. Category filtering, quantity totals, whole-line removal and empty checkout work without errors. Checkout clears the cart and claims an order was placed despite being only local demo logic. Text-only cards, unlabeled filter state and non-live toast feedback limit finish.", "reviewInputBytes": 10864, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/0ab63f0f62.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/postmortem.md", "nativeTokensPerSec": 12.82, "nativeGeneratedTokens": 11233, "promptTokensPerSec": 676.03, "usageSource": "engine-log", "durationMs": 1282918, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3114013671875 }, "peakRssMb": 19305, "peakGpuMemMb": 10628, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-petshop", "model": "qwen3.8-27b-q2", "scenario": "petshop", "repetition": 2, "trialId": "petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-20T23:12:54.590Z", "finishedAt": "2026-10-21T00:55:59.250Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"pet-shop-website/workspace/index.html:3:targetnone:frnone:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 6 tool calls (0 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 3.7, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 3, "codeQuality": 5, "polish": 3 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent layout and working filters, sorting, cart arithmetic and form validation. Required PNG is absent. Mobile menu opens above the viewport; fresh-page replay confirms inaccessible links. Reservation/contact flows falsely claim delivery while only changing local state. Phone link differs from displayed number and today's hours are replaced by vague status.", "reviewInputBytes": 19977, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Only the19977-byte documented source selection was read; the remaining9879CSS characters were omitted, so code/style assessment is partial. Full HTML structure and JavaScript were included.", "Full unchanged page rendered offline at1200x900 and390x844 with desktop Chromium, not mobile-device emulation; external HTTP(S) blocked.", "Targeted controls and validation were replayed. No messages, reservations, calls or emails were sent externally. Successful local status text does not establish delivery.", "Initial collector stopped on an offscreen mobile navigation click. Second replay recorded the failure, and a fresh390px context at scroll0 confirmed it; see5e42f6e4ee-mobile-fresh.json.", "Screenshots at full page height are scaled in review; viewport screenshots inspected for legible desktop, cart and mobile top/menu states. No exhaustive accessibility or game-tree-style interaction audit." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/postmortem.md", "nativeTokensPerSec": 4.48, "nativeGeneratedTokens": 22816, "promptTokensPerSec": 177.83, "usageSource": "engine-log", "durationMs": 6184660, "toolCalls": 26, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3480224609375 }, "peakRssMb": 18804, "peakGpuMemMb": 11610, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-petshop", "model": "gemma4-12b-q4", "scenario": "petshop", "repetition": 2, "trialId": "petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T00:57:41.805Z", "finishedAt": "2026-10-21T01:51:11.176Z", "success": false, "failureClass": "model", "reason": "repair-aborted: eneko/cbc5a593 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.", "modelEligible": true, "fixedComposite": 3.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 3, "codeQuality": 4, "polish": 2 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable product/service cards and labeled booking fields fit both widths. Required PNG and logo img are absent; only a comment remains. Cart buttons merely alert. Native required-field validation works, but booking claims delivery and a callback without any contact field or submission. External hero photo was blocked in offline review; its live appearance is unassessed.", "reviewInputBytes": 7992, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected textual review input is within 20 KiB. Images assessed as rendered in the final page.", "Rendered at 1200x900 and 390x844 with desktop Chromium; this is not mobile device emulation.", "External HTTP(S) requests were blocked for consistent offline rendering. External imagery availability and appearance were not assessed.", "Interaction checks are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bf0a2b6d90.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/postmortem.md", "nativeTokensPerSec": 11.11, "nativeGeneratedTokens": 23271, "promptTokensPerSec": 423.27, "usageSource": "engine-log", "durationMs": 3209371, "toolCalls": 42, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.377288818359375 }, "peakRssMb": 18061, "peakGpuMemMb": 8616, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-petshop", "model": "gemma4-e4b-q4", "scenario": "petshop", "repetition": 2, "trialId": "petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T01:52:41.109Z", "finishedAt": "2026-10-21T05:02:42.689Z", "success": false, "failureClass": "model", "reason": "repair-aborted: baboucarr/bd2eb08d 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.", "modelEligible": true, "fixedComposite": 1.3, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 3 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive three-product page includes the generated PNG and runs without errors. Cart buttons explicitly disclose simulated behavior. Readable, simple source meets the page/logo brief, but default controls, sparse content and garbled logo lettering limit polish. The page was left in Default; its official orchestration failure is separate from this artifact assessment.", "reviewInputBytes": 2724, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML source (2724 bytes) was read within the combined 20 KiB cap. Unused SVG and unrelated coordination/review documents were excluded from artifact judging.", "Generated PNG was assessed as rendered in the final page and separately hashed; it was not treated as textual source input.", "Rendered at 1200x900 and 390x844 in offline desktop Chromium; this is not mobile-device emulation. Three cart-button interactions and layout checks were replayed without errors or horizontal overflow.", "This is the preserved Default-workspace output. No dedicated work project was created, so the frozen evaluator did not consider this page. Artifact scores do not change the official gate or failure classification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/580190de45.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/postmortem.md", "nativeTokensPerSec": 19.51, "nativeGeneratedTokens": 192186, "promptTokensPerSec": 714.51, "usageSource": "engine-log", "durationMs": 11401580, "toolCalls": 240, "redFlags": [ { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/sessions/baboucarr", "tool": "(no write_file this turn)", "argsSummary": "content_chars=3750 h2_count=4 preview=\"# Release Readiness Data Gathering — analysis ## Summary I have compiled an initial assessment of the data required by …\"", "atTurn": 1, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." }, { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/sessions/baboucarr", "tool": "(no write_file this turn)", "argsSummary": "content_chars=3824 h2_count=3 preview=\"# Plan to Acquire Missing External Evidence for Dimitrios Since direct system access (JIRA, CI servers, Datadog/Prometh…\"", "atTurn": 10, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." }, { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/sessions/claudia", "tool": "(no write_file this turn)", "argsSummary": "content_chars=1125 h2_count=2 preview=\"I've completed the structural review of `index.html` and compiled the findings into a report for you. Here is the summa…\"", "atTurn": 4, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." }, { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/sessions/claudia", "tool": "(no write_file this turn)", "argsSummary": "content_chars=1124 h2_count=2 preview=\"The structural review of `index.html` is complete, and I have documented my findings in `report.md`. Here is the summar…\"", "atTurn": 7, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." } ], "autoAnswers": 6, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 9, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.563812255859375 }, "peakRssMb": 14263, "peakGpuMemMb": 6870, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-tankcombat", "model": "ornith1.5-9b-q4", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T05:04:50.385Z", "finishedAt": "2026-10-21T05:27:57.555Z", "success": true, "failureClass": "pass", "reason": "this-work-e/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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 3, "codeQuality": 3, "polish": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Attractive HUD and overlays, but Arrow/Space leave the opening screen inert; only R starts. It then auto-fires regardless of Space. Movement, pause, natural damage/game-over and restart work. Frame-based motion and fixed 800px canvas weaken robustness; narrow layout clips the arena and score. No scored kill was observed.", "reviewInputBytes": 17102, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML source (17102 bytes) was read within20KiB. No other candidate source was read.", "Ordinary keyboard input and read-only game-state inspection; no injected entities, scores, timers or deterministic random override. Scoring logic was inspected in source, but no scored kill was observed in the short random replay.", "Offline desktop Chromium1200x900 and390x844 screenshots inspected; narrow viewport is not mobile-device emulation. Targeted control checks are not an exhaustive playtest.", "The first review-only collector attempt failed from a transpiled __name reference before collecting evidence; the corrected collector completed. This was not a candidate exception." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/b8c3abcaae.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/postmortem.md", "nativeTokensPerSec": 9.61, "nativeGeneratedTokens": 11287, "promptTokensPerSec": 484.86, "usageSource": "engine-log", "durationMs": 1387170, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3768463134765625 }, "peakRssMb": 9975, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-tankcombat", "model": "qwen3.5-4b-q4", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T05:29:22.702Z", "finishedAt": "2026-10-21T05:30:35.895Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 3, "codeQuality": 2, "polish": 2 }, "advisoryMean": 2.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Movement and firing work, with natural collision loss and restart. Bullets travel 90 degrees from the barrel; cooldown units cause near-frame-rate firing. Source sends enemy shots away and tests them against enemies, never the player. No visible instructions or viewport metadata; fixed canvas clips at390px. No scored kill was observed.", "reviewInputBytes": 10721, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML source (10721 bytes) was read within20KiB. No other candidate source was read.", "Ordinary keyboard input and read-only exposed state/lexical player inspection; no injected game state. Replay confirms movement, downward projectiles after rightward orientation, natural game-over and restart. Enemy projectile ownership defects are source findings; no scored kill was observed.", "Offline desktop Chromium1200x900 and390x844 screenshots inspected; narrow viewport is not mobile-device emulation. Short random gameplay is not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bd466d8610.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/postmortem.md", "nativeTokensPerSec": 122.72, "nativeGeneratedTokens": 4773, "promptTokensPerSec": 3742.62, "usageSource": "engine-log", "durationMs": 73193, "toolCalls": 4, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2860260009765625 }, "peakRssMb": 6583, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-tankcombat", "model": "underdog-saluki-27b-iq2-mix", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T05:32:57.533Z", "finishedAt": "2026-10-21T05:39:12.859Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 2, "codeQuality": 3, "polish": 1 }, "advisoryMean": 2, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Arrow movement, enemy fire, lives, natural game-over and R restart work. Space never fires and WASD does not move: a code-key map is indexed with event.key. Bare rectangles, detached HUD and no visible control instructions; fixed canvas overflows390px. Combat is incomplete because the player cannot shoot with the requested key.", "reviewInputBytes": 7813, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML source (7813 bytes) was read within20KiB. No other candidate source was read.", "Ordinary ArrowRight, d, Space and r input with read-only game-state snapshots. No source edits, synthetic KeyboardEvent values or injected game state. Short replay observed natural damage and game-over; scoring logic was read but no scored kill occurred.", "Offline desktop Chromium1200x900 and390x844 screenshots inspected; narrow viewport is not mobile-device emulation. No exhaustive gameplay or touch-device test." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/e0ce0d6908.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/postmortem.md", "nativeTokensPerSec": 13.17, "nativeGeneratedTokens": 3663, "promptTokensPerSec": 689.75, "usageSource": "engine-log", "durationMs": 375326, "toolCalls": 4, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2756195068359375 }, "peakRssMb": 14521, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-tankcombat", "model": "qwen3.8-27b-q2", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T05:40:42.477Z", "finishedAt": "2026-10-21T06:09:01.453Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 2 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable arena, HUD and controls. Arrow/WASD movement and Space projectiles respond correctly; enemy fire causes a clear game-over. Source includes score updates and enemy replenishment, though no scored kill was observed. No restart control, sparse feedback and a fixed canvas that clips on narrow screens limit polish.", "reviewInputBytes": 7852, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML source (7852 bytes) was read within20KiB. No other candidate source was read.", "Ordinary keyboard input and read-only game-state inspection; no injected state, fake timers or random override. Natural enemy damage/game-over was observed. Score/replenishment logic was inspected in source; no scored kill was replayed.", "Offline desktop Chromium1200x900 and390x844 screenshots inspected; narrow viewport is not mobile-device emulation. Short random replay does not establish exhaustive gameplay correctness." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/036b15091a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/postmortem.md", "nativeTokensPerSec": 3.94, "nativeGeneratedTokens": 6173, "promptTokensPerSec": 325.63, "usageSource": "engine-log", "durationMs": 1698976, "toolCalls": 2, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.291656494140625 }, "peakRssMb": 15015, "peakGpuMemMb": 11604, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-tankcombat", "model": "gemma4-12b-q4", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T06:10:55.185Z", "finishedAt": "2026-10-21T06:19:01.267Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 5, "codeQuality": 3, "polish": 2 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive canvas, turn-and-throttle movement and directionally consistent Space shots work. Enemies pursue the player; natural collision triggers a game-over alert and automatic reload. Sparse rectangles make heading unclear. Frame-based simulation and array splicing during forEach weaken robustness. Scoring is implemented but no scored kill was observed.", "reviewInputBytes": 7759, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML source (7759 bytes) was read within20KiB. No other candidate source was read.", "ArrowRight and d rotate this tank rather than translate it; ArrowUp was replayed separately to verify forward movement. Ordinary Space input produced projectiles. Natural collision displayed Game Over and the candidate reloaded after the review browser dismissed the alert.", "Read-only state inspection and ordinary keyboard input only. No candidate edits, injected entities/state or fake random/timing. Scoring was inspected in source but no scored kill occurred during the short random replay.", "Desktop1200x900 and narrow390x844 screenshots inspected offline; the canvas fits both. Narrow desktop viewport is not mobile-device emulation; no touch-control or exhaustive collision audit." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1d7324a5a6.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/postmortem.md", "nativeTokensPerSec": 13.56, "nativeGeneratedTokens": 5273, "promptTokensPerSec": 433.29, "usageSource": "engine-log", "durationMs": 486082, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2980194091796875 }, "peakRssMb": 13451, "peakGpuMemMb": 8616, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-tankcombat", "model": "gemma4-e4b-q4", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T06:20:45.369Z", "finishedAt": "2026-10-21T06:25:39.787Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 3, "codeQuality": 2, "polish": 1 }, "advisoryMean": 2.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Rotation, forward thrust, Space shots and natural game-over/reload work. Down does nothing; rendered heading and shot/thrust direction disagree vertically. Thrust is excessive and frame-dependent. Enemy health is not reset on respawn. Sparse boxes, no instructions or player-health display, and a clipped fixed canvas limit usability. No scored kill was observed.", "reviewInputBytes": 9993, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML9993bytes read within20KiB. Ordinary keyboard input and read-only lexical game-state snapshots only; no injected state, time or random values.", "Initial300ms firing sample had no remaining bullet after forward thrust reached the top edge. A separate fresh-page50ms probe confirmed a Space projectile; see advisory-browser-evidence/1b5d0b9657-firing-direction.json and its screenshot. This avoids mistaking an expired projectile for broken firing.", "Natural game-over alerts and candidate reload occurred; the review browser dismissed the alerts. Scoring and missing enemy-health reset were inspected in source, but no scored kill/respawn was replayed.", "Desktop1200x900 and narrow390x844 screenshots inspected offline. Narrow viewport is not mobile-device emulation; short gameplay does not establish exhaustive correctness." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1b5d0b9657.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/postmortem.md", "nativeTokensPerSec": 26.19, "nativeGeneratedTokens": 5570, "promptTokensPerSec": 987.3, "usageSource": "engine-log", "durationMs": 294418, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4006805419921875 }, "peakRssMb": 6235, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-tankcombat", "model": "qwen3.5-9b-q4", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T06:27:15.603Z", "finishedAt": "2026-10-21T06:38:55.897Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 4, "codeQuality": 3, "polish": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Arrow/WASD movement and upward Space shots work, with pursuing enemies, score and collision particles. Holding Down moves the player offscreen to y785 in a600px arena: the vertical clamp uses width. Frame-based updates and an input counter that advances while idle weaken validation. Fixed canvas clips narrow screens; no restart. No scored kill observed.", "reviewInputBytes": 6822, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full final HTML6822UTF8bytes read within20KiB; frozen sniff reports6820characters. No other candidate source read.", "Ordinary keyboard input and read-only lexical state only. ArrowRight and d each moved64pixels, Space created a projectile, and Down reachedy785 while canvas.height=600. No injected collisions/entities/state or modified clock/randomness.", "Input tick advanced11to29 without input. Scoring/collision code was inspected; random short replay observed no scored kill, so score correctness is not claimed from playback.", "Desktop1200x900 and narrow390x844 screenshots inspected offline. Canvas remains804px wide despite document.scrollWidth=390 because overflow is hidden; screenshot/state establish clipping. Narrow viewport is not mobile-device emulation." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/postmortem.md", "nativeTokensPerSec": 9.52, "nativeGeneratedTokens": 4120, "promptTokensPerSec": 467.49, "usageSource": "engine-log", "durationMs": 700294, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.383056640625 }, "peakRssMb": 12418, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-schema-migration", "model": "qwen3.5-4b-q4", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T06:40:18.887Z", "finishedAt": "2026-10-21T06:47:47.258Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Types, store and handlers preserve public APIs; six name cases split correctly without input mutation. Required tests and guide exist, and strict source/consumer checks work. Migration copies only id/email and drops extra fields despite the guide's preservation claim. Tests miss that loss; stale comments and uneven formatting reduce clarity.", "reviewInputBytes": 10244, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected textual input plus original fixture grounding totals10244bytes, below20KiB. Shared/default project metadata was not assessed.", "Production exports were executed with six ordinary name inputs and extra role/active/metadata fields. All seeded id/email fields survive; arbitrary added fields are lost. The latter exceeds the seeded deterministic probe and tests the broader brief/documentation promise, not a replacement official verdict.", "Store add/get/list and all three handlers were exercised. Strict source compilation had no errors; a virtual consumer accepts normal CreateUserInput and rejects stale User.name. Model-authored Vitest tests were inspected, not executed.", "No candidate files or frozen gates changed. These targeted probes are not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/0ab542d5de.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/postmortem.md", "nativeTokensPerSec": 16.09, "nativeGeneratedTokens": 4001, "promptTokensPerSec": 689.03, "usageSource": "engine-log", "durationMs": 448371, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.416717529296875 }, "peakRssMb": 9489, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-schema-migration", "model": "underdog-saluki-27b-iq2-mix", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T06:48:39.813Z", "finishedAt": "2026-10-21T07:12:05.773Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear, focused refactor preserves store/handler APIs, splits six name cases correctly, and supplies three tests plus a migration guide. Strict source typing works and stale User.name is rejected. Migration drops all extra record fields, contradicting the guide's explicit future-field claim; tests only verify seeded id/email preservation. Old comments remain.", "reviewInputBytes": 9586, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected textual input plus original fixture grounding totals9586bytes, below20KiB. Shared/default project metadata was not assessed.", "Production exports were executed with six ordinary name inputs and extra role/active/metadata fields. All seeded id/email fields survive; arbitrary added fields are lost. The latter exceeds the seeded deterministic probe and tests the broader brief/documentation promise, not a replacement official verdict.", "Store add/get/list and all three handlers were exercised. Strict source compilation had no errors; a virtual consumer accepts normal CreateUserInput and rejects stale User.name. Model-authored Vitest tests were inspected, not executed.", "No candidate files or frozen gates changed. These targeted probes are not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/0c366c4af3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/postmortem.md", "nativeTokensPerSec": 3.06, "nativeGeneratedTokens": 2569, "promptTokensPerSec": 144.06, "usageSource": "engine-log", "durationMs": 1405960, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3477935791015625 }, "peakRssMb": 17774, "peakGpuMemMb": 10628, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-schema-migration", "model": "qwen3.8-27b-q2", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T07:13:26.235Z", "finishedAt": "2026-10-21T07:33:31.266Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clean, focused refactor preserves extra fields, removes name without mutation and correctly splits six cases, including whitespace edges. Store and handler APIs work; strict typing rejects stale User.name. Three meaningful tests and an accurate migration guide meet the brief. Its own tests do not cover extra-field preservation; the old README remains stale.", "reviewInputBytes": 10268, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected textual input plus original fixture grounding totals10268bytes, below20KiB. Review is unblinded; no unrelated shared/default artifacts assessed.", "Production exports were executed with six ordinary name inputs and extra role/active/metadata fields. Both seeded id/email and added fields survived; name was removed and the input remained unchanged. These probes are targeted, not exhaustive.", "Store add/get/list and all three handlers were exercised. Strict source compilation had no errors; a virtual consumer accepts normal CreateUserInput and rejects stale User.name. Candidate Vitest tests were inspected, not executed.", "Extra-field and whitespace probes extend beyond the frozen three-case deterministic runtime check. No candidate files, official scores, classifications or gates changed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/45f9a0afd5.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/postmortem.md", "nativeTokensPerSec": 4.44, "nativeGeneratedTokens": 2617, "promptTokensPerSec": 116.04, "usageSource": "engine-log", "durationMs": 1205031, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.330322265625 }, "peakRssMb": 18366, "peakGpuMemMb": 11608, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-schema-migration", "model": "gemma4-12b-q4", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T07:34:37.946Z", "finishedAt": "2026-10-21T07:45:34.572Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Straightforward refactor preserves public APIs, splits all six name cases and supplies three full-record tests plus a detailed guide. Source and consumer typing pass. Migration copies only id/email and drops extra fields despite broader preservation claims. Tests omit that loss, the guide overstates verification, and old comments still describe User.name.", "reviewInputBytes": 9872, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected textual input plus original fixture grounding totals9872bytes, below20KiB. Review is unblinded; no unrelated shared/default artifacts assessed.", "Six ordinary name cases pass without input mutation or retained name. Seeded id/email survive, but added role/active/metadata fields are lost. This tests the broader brief/documentation claim and does not replace the frozen verdict.", "Store add/get/list and all three handlers were exercised. Strict source compilation had no errors; a virtual consumer accepts normal CreateUserInput and rejects stale User.name. Candidate tests were inspected, not executed; their source verifies only the listed fields.", "No candidate or gate changes. Targeted probes are not an exhaustive behavior audit." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/c9a0db57f3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/postmortem.md", "nativeTokensPerSec": 12.01, "nativeGeneratedTokens": 3628, "promptTokensPerSec": 430.47, "usageSource": "engine-log", "durationMs": 656626, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4029083251953125 }, "peakRssMb": 16751, "peakGpuMemMb": 8620, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-schema-migration", "model": "gemma4-e4b-q4", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T07:46:38.923Z", "finishedAt": "2026-10-21T07:55:13.072Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 4 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Store, handlers and strict types work, with four tests and a migration guide. Migration handles ordinary names but trims/collapses whitespace instead of splitting at the first space, failing two edge probes. Extra fields are lost despite preservation claims. Its added whitespace test endorses the changed semantics; first-name sorting also switches to binary comparison.", "reviewInputBytes": 10815, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected textual input plus original fixture grounding totals10815bytes, below20KiB. Review is unblinded; no unrelated shared/default artifacts assessed.", "Four of six name probes match literal first-space splitting. Leading-space and repeated-space cases fail because trim and a whitespace regex normalize the input. Input is not mutated and name is removed; seeded id/email survive, extra role/active/metadata fields do not.", "Public store/handler probes and strict source compilation pass; a virtual consumer accepts normal CreateUserInput and rejects stale User.name. Candidate Vitest tests were read, not executed. The fourth test explicitly expects normalization contrary to literal first-space behavior.", "The frozen runtime gate covers three ordinary names and seeded non-name fields; the advisory whitespace/extra-field probes are distinct. No official results or candidate files changed; sorting change is a source finding, not a measured cross-locale failure." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/5290a21937.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/postmortem.md", "nativeTokensPerSec": 23.29, "nativeGeneratedTokens": 5571, "promptTokensPerSec": 1015.96, "usageSource": "engine-log", "durationMs": 514149, "toolCalls": 18, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.39794921875 }, "peakRssMb": 6305, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-schema-migration", "model": "qwen3.5-9b-q4", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T07:56:04.460Z", "finishedAt": "2026-10-21T08:08:28.239Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 3, "scopeDiscipline": 4 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Source handles all six name cases, preserves public APIs and passes strict type probes; three required tests exist. Migration drops extra fields. The guide invents optional CreateUserInput.id in both schemas, while the actual contract has none. Sorting becomes case-insensitive without a requirement, and stale comments remain. Tests miss metadata loss.", "reviewInputBytes": 9388, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected textual input plus original fixture grounding totals9388bytes, below20KiB. Review is unblinded; no unrelated shared/default artifacts assessed.", "Six ordinary name cases pass without mutation or retained name. Seeded id/email survive, but added role/active/metadata fields are lost. The latter extends beyond the frozen seeded-field probe and is not a replacement verdict.", "Store add/get/list and all three handlers were exercised. Strict source compilation had no errors; a virtual consumer accepts normal CreateUserInput and rejects stale User.name. Candidate Vitest tests were inspected, not executed.", "The documentation's optional id and new case-insensitive sorting were identified from full source and original fixtures. No locale matrix or exhaustive sort test was run; no candidate files or gates changed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/29bc00c4a6.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/postmortem.md", "nativeTokensPerSec": 10.55, "nativeGeneratedTokens": 4300, "promptTokensPerSec": 477.23, "usageSource": "engine-log", "durationMs": 743779, "toolCalls": 31, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.499176025390625 }, "peakRssMb": 15105, "peakGpuMemMb": 8104, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-schema-migration", "model": "ornith1.5-9b-q4", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T08:09:19.817Z", "finishedAt": "2026-10-21T08:39:18.182Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear refactor preserves store/handler APIs and splits all six name cases without mutation. Strict source typing works and stale User.name is rejected. Required tests and a concise guide exist. Migration copies only id/email and drops every extra record field; tests miss that loss. The multi-word test also omits its id assertion, and an old handler comment remains.", "reviewInputBytes": 8081, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected text and original fixture grounding total8081bytes, below20KiB; review is unblinded.", "Six name cases and added role/active/metadata fields were probed through the actual production export. Seeded id/email survive and the input is unchanged; added fields are lost. The metadata check exceeds the frozen seeded-field gate and does not replace its verdict.", "Store add/get/list, three handlers and strict source/virtual-consumer compilation were checked. Candidate Vitest tests were inspected, not executed. Probes are targeted, not exhaustive; no candidate files changed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/58ec2d685e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/postmortem.md", "nativeTokensPerSec": 4.9, "nativeGeneratedTokens": 7200, "promptTokensPerSec": 560.58, "usageSource": "engine-log", "durationMs": 1798365, "toolCalls": 32, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4395751953125 }, "peakRssMb": 14797, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec", "model": "underdog-saluki-27b-iq2-mix", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T08:40:19.115Z", "finishedAt": "2026-10-21T08:46:48.853Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Concise transition table and shared guard logic pass all84 ordinary state/event/option probes, including exact errors and history. Strict typing passes and seeded tests/configs are unchanged. Inherited keys such as constructor, toString and __proto__ are accepted as events in all36 extra probes, corrupting state to undefined instead of rejecting them.", "reviewInputBytes": 7117, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected source and grounding input totals7117bytes, below20KiB; review is unblinded.", "Only the production factory was executed. Six states are reached by legal sends; two option settings and ten string events yield84 ordinary plus36 inherited-name probes. Default construction checked separately. No state injection or candidate edits.", "Ordinary results include can/send agreement, exact rejection strings, nonmutation on rejection and ordered history. Extra object-property names extend the frozen test examples; all36 were actually reachable and fail.", "Seeded tests/configs are byte-identical to the original fixtures. Existing official Vitest/tsc outcomes are preserved; the advisory collector adds production probes and strict source compilation, not a replacement verdict or exhaustive API audit." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/984f57eb76.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/postmortem.md", "nativeTokensPerSec": 3.19, "nativeGeneratedTokens": 615, "promptTokensPerSec": 152.28, "usageSource": "engine-log", "durationMs": 389738, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2672882080078125 }, "peakRssMb": 11778, "peakGpuMemMb": 10624, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-failing-tests-spec", "model": "qwen3.8-27b-q2", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T08:47:52.031Z", "finishedAt": "2026-10-21T09:04:51.873Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Well-structured typed transitions and shared guard handling pass84 ordinary lifecycle/history/error probes. Strict source typing passes and the seeded specification is intact. Casting arbitrary strings to Event does not validate them: inherited object-property names are accepted in all36 extra probes and set state to undefined.", "reviewInputBytes": 7213, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected source and grounding input totals7213bytes, below20KiB; review is unblinded.", "Only the production factory was executed, reaching states by legal sends. Two option settings, six states and ten string events produce84 ordinary and36 inherited-name probes; default construction is separate. All special-name states were reachable.", "Strict source compilation and exact runtime state/history/error behavior were checked. Candidate source/tests/configs were not edited. The unchanged official test run is not replaced by these advisory probes.", "Tests cover targeted finite cases, not every possible string or caller mutation. No external dependency or model was installed or invoked." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/7207f1d1ce.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/postmortem.md", "nativeTokensPerSec": 5.02, "nativeGeneratedTokens": 1386, "promptTokensPerSec": 127.65, "usageSource": "engine-log", "durationMs": 1019842, "toolCalls": 12, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.361419677734375 }, "peakRssMb": 18453, "peakGpuMemMb": 11606, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-failing-tests-spec", "model": "gemma4-12b-q4", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T09:05:55.056Z", "finishedAt": "2026-10-21T09:07:51.810Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable implementation passes all84 ordinary lifecycle, guard, history and exact-error probes; strict typing and seed integrity pass. Validation is duplicated between can/send. Using in admits inherited event names: all36 extra probes corrupt state to a function or object instead of rejecting the unknown string.", "reviewInputBytes": 7556, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected source and grounding input totals7556bytes, below20KiB; review is unblinded.", "Only the production factory was executed. Six legally reached states, two option settings and ten events yield84 ordinary and36 inherited-name probes. All extra cases were reachable; ordinary can/send/history/error behavior is checked.", "Strict source compilation and unchanged seed hashes were recorded. No candidate edits or injected state. The probes are advisory and do not change the existing official Vitest/tsc PASS.", "No exhaustive arbitrary-string, concurrency or caller-mutation audit was performed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/1e1ef96120.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/postmortem.md", "nativeTokensPerSec": 13.64, "nativeGeneratedTokens": 645, "promptTokensPerSec": 440.22, "usageSource": "engine-log", "durationMs": 116754, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.256134033203125 }, "peakRssMb": 11505, "peakGpuMemMb": 8616, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-failing-tests-spec", "model": "gemma4-e4b-q4", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T09:08:44.903Z", "finishedAt": "2026-10-21T09:16:26.733Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on src/machine.ts; missing signals stayed unchanged: vitest-green, tsc-clean. Last failure: vitest run failed (exit 1): × submit moves draft to pending and returns the new state 2ms | × pay moves pending to paid 0ms | × ship moves paid to shipped, deliver moves shipped to delivered 0ms | × cancel is allowed from draft, pending, and paid 0ms | FAIL tests/machine.test.ts > order machine: transitions > submit moves draft to pending and returns the new state | TypeError: Assignment to constant variable. | FAIL tests/machine.test.ts > order machine: transitions > pay moves pending to paid | FAIL tests/machine.test.ts > order machine: transitions > ship moves paid to shipped, deliver moves shipped to delivered || tsc --noEmit failed (exit 2): src/machine.ts(39,9): error TS2322: Type '(op", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 1, "codeQuality": 2, "robustness": 1, "scopeDiscipline": 3 }, "advisoryMean": 1.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Initial construction works, but submit/cancel throw Assignment to constant variable: send shadows state with const and then assigns it. Later states are unreachable. Source also exposes a static state value, ignores guards in can, and broadens send to void. Strict compilation reports the const assignment and optional-boolean guard. Seeded tests/configs remain intact.", "reviewInputBytes": 7756, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected source and grounding input totals7756bytes, below20KiB; review is unblinded.", "Default construction and initial-state ordinary API calls were executed without candidate edits or injected state. Submit and cancel throw; all five other states cannot be reached by legal sends.", "Collector totals10/84 ordinary and0/36 special include70 ordinary and30 special setup failures. Only14 ordinary and6 special draft cases directly executed; counts must not imply120 independent reachable behaviors were tested.", "Static state exposure, guard omission in can and changed send return type are source findings. Strict compiler diagnostics confirm optional-boolean guard code2322 and const assignment code2588. No repair was applied; frozen repair-exhausted/model failure and fixed score remain unchanged." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/aa491ce09f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/postmortem.md", "nativeTokensPerSec": 23.97, "nativeGeneratedTokens": 5017, "promptTokensPerSec": 941.75, "usageSource": "engine-log", "durationMs": 461830, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3009796142578125 }, "peakRssMb": 7706, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-failing-tests-spec", "model": "qwen3.5-9b-q4", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T09:17:21.473Z", "finishedAt": "2026-10-21T09:23:51.671Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear typed transition table and shared validation pass all 84 ordinary lifecycle, guard, error and history probes. Strict typing passes and seeded tests/configs are unchanged. Arbitrary strings cast to Event still reach inherited object properties: all 36 special-name probes accept invalid events and corrupt state to a function or object.", "reviewInputBytes": 7049, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected source and grounding total 7049 bytes, below 20KiB. Review is unblinded.", "Production factory only; all six states reached with legal sends. Two option settings and ten string events yield 84 ordinary and 36 inherited-name probes. Default construction separately checked.", "Strict source compilation and seed equality checked. Advisory probes supplement, without changing, frozen test outcomes; targeted cases are not exhaustive. No candidate edits or official Vitest rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9a48f2a8dc.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/postmortem.md", "nativeTokensPerSec": 11.62, "nativeGeneratedTokens": 1563, "promptTokensPerSec": 517.85, "usageSource": "engine-log", "durationMs": 390198, "toolCalls": 15, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.419921875 }, "peakRssMb": 12110, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-failing-tests-spec", "model": "ornith1.5-9b-q4", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T09:24:44.819Z", "finishedAt": "2026-10-21T09:46:37.579Z", "success": false, "failureClass": "model", "reason": "no real progress for 721s — model busy but not delivering (hard digest stuck at 7a72d63aab99da16)", "modelEligible": true, "fixedComposite": 5.6, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 5 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Compact code updates its closure and history, but returns state as a stale value: submit returns pending while public state remains draft. The main observable contract fails. Inherited event names also bypass validation. Strict typing passes and seeded tests/configs remain intact; readable source does not compensate for broken public state.", "reviewInputBytes": 6723, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected source and grounding total 6723 bytes, below 20KiB; unblinded review.", "Production factory only, with legal setup sends and no state injection. Only 14 ordinary draft cases execute: 10 pass and four valid transitions fail the public-state contract. Seventy other ordinary cases fail setup because the public state stays draft.", "Six inherited-name draft cases execute and fail; 30 special-name cases fail setup. Totals 10/84 and 0/36 include setup failures, not 120 independently reachable tests. Strict source diagnostics empty; seeds unchanged.", "Internal progression is visible in source and existing official test results, but the collector does not bypass stale public state to claim full runtime coverage. No candidate edits or gate reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/ea1c1d9f7e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/postmortem.md", "nativeTokensPerSec": 4.98, "nativeGeneratedTokens": 1807, "promptTokensPerSec": 539.69, "usageSource": "engine-log", "durationMs": 1312760, "toolCalls": 18, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2982025146484375 }, "peakRssMb": 12654, "peakGpuMemMb": 7256, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-failing-tests-spec", "model": "qwen3.5-4b-q4", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T09:47:38.571Z", "finishedAt": "2026-10-21T09:57:04.648Z", "success": false, "failureClass": "model", "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 2 while the failing detail kept changing; latest checked file src/machine.ts. Last failure: vitest run failed (exit 1): × rejects cancel after shipping 2ms | FAIL tests/machine.test.ts > order machine: invalid transitions > rejects cancel after shipping | AssertionError: expected [Function] to throw an error | - Expected: | + Received: || tsc --noEmit failed (exit 2): src/machine.ts(91,20): error TS2454: Variable 'toState' is used before being assigned.", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 3, "robustness": 3, "scopeDiscipline": 3 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Most lifecycle behavior works: 82/84 ordinary probes and all 36 inherited-name rejections pass. Shipping incorrectly remains cancellable in both guard modes. Strict typing rejects possibly unassigned toState. Validation is duplicated and the public state/history lose readonly declarations. Seeded tests and configuration are unchanged.", "reviewInputBytes": 8070, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete source and grounding total 8070 bytes, below 20KiB; unblinded review.", "All six states are reachable through legal sends. Two option settings and ten events exercise can/send agreement, exact errors, rejection nonmutation and ordered history. Both shipped-cancel cases fail; all 36 inherited-name cases reject correctly.", "Strict source compiler reports TS2454 at line 91. The probe executes existing production exports via TypeScript transpilation, not a repaired copy. Seed equality verified; no candidate changes or official test rerun.", "Readonly declaration loss is a source/API finding; arbitrary external state mutation was not exercised. Targeted finite cases do not exhaust the public interface." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/f10301a18a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/postmortem.md", "nativeTokensPerSec": 17.07, "nativeGeneratedTokens": 5418, "promptTokensPerSec": 746.7, "usageSource": "engine-log", "durationMs": 566077, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3394775390625 }, "peakRssMb": 11703, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-symptom-debug", "model": "qwen3.8-27b-q2", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T09:57:57.495Z", "finishedAt": "2026-10-21T10:06:24.229Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Minimal cursor fix preserves every record across varied page sizes and updates the comment to match the next-index contract. All 62 targeted pagination, invalid-input and direct-cursor checks pass. Acceptance script and package remain unchanged; the correction stays within scope.", "reviewInputBytes": 4169, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected text and original grounding total 4169 bytes, below 20KiB; unblinded review.", "Only actual production exports executed: 48 varied-size/limit full walks, six invalid-limit checks, three malformed-cursor checks and five direct-index cases. Walks also verify input nonmutation.", "Seeded acceptance script/package byte equality checked. No official acceptance rerun, candidate edits or new dependencies. These 62 checks are targeted, not exhaustive cursor-format or malformed-type coverage; permissive seed parsing remains outside this minimal correction." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/aa24edd324.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/postmortem.md", "nativeTokensPerSec": 3.24, "nativeGeneratedTokens": 919, "promptTokensPerSec": 114.26, "usageSource": "engine-log", "durationMs": 506734, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2337799072265625 }, "peakRssMb": 11792, "peakGpuMemMb": 11602, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-symptom-debug", "model": "gemma4-12b-q4", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T10:07:26.141Z", "finishedAt": "2026-10-21T10:18:42.506Z", "success": false, "failureClass": "model", "reason": "no real progress for 602s — model busy but not delivering (hard digest stuck at 52b0c563f1a3b9f5)", "modelEligible": true, "fixedComposite": 3.8, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-code-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "No artifact", "justification": "The final pagination module is byte-identical to the buggy seed; no produced correction exists. Acceptance script and package are unchanged. Two successful reads and one rejected installed-script call precede the recorded no-progress stop. Fixture code receives no advisory scores.", "reviewInputBytes": 4155, "protocolDeviation": false, "reviewLimitations": [ "No model-produced correction exists; unchanged fixture code is not an output. No quality axes or mean assigned.", "Seed equality and full workspace inventory verified. Raw tool events confirm reads and the rejected script call; native logs show generation continued until harness shutdown. The incomplete persisted session is not treated as proof of no activity.", "Frozen model-stuck classification is preserved. No dependency failure, GPU crash or retry has been inferred from an absent output.", "Model identity was known during review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/7de98bb335.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/postmortem.md", "nativeTokensPerSec": 11.65, "nativeGeneratedTokens": 235, "promptTokensPerSec": 425.81, "usageSource": "engine-log", "durationMs": 676365, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.260345458984375 }, "peakRssMb": 11366, "peakGpuMemMb": 8616, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-symptom-debug", "model": "gemma4-e4b-q4", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T10:19:35.223Z", "finishedAt": "2026-10-21T10:21:15.593Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The focused one-line fix removes the extra cursor increment and passes all 62 targeted pagination, invalid-input and direct-cursor checks. Acceptance script and package remain unchanged. The adjacent comment still incorrectly says to resume one past the cursor, contradicting the corrected behavior.", "reviewInputBytes": 4151, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Complete selected text totals 4151 bytes, below 20KiB; review is unblinded.", "Actual production exports exercised through 48 varied-size/limit walks, six invalid-limit checks, three malformed-cursor checks and five direct-index cases. Walks verify input nonmutation.", "Acceptance script/package remain byte-identical. No candidate edits or frozen acceptance rerun. Targeted checks are not exhaustive; inherited permissive cursor parsing was not rewritten or broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9f5f1437ef.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/postmortem.md", "nativeTokensPerSec": 23.41, "nativeGeneratedTokens": 1525, "promptTokensPerSec": 1077.08, "usageSource": "engine-log", "durationMs": 100370, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.247833251953125 }, "peakRssMb": 4488, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-symptom-debug", "model": "qwen3.5-9b-q4", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T10:22:10.029Z", "finishedAt": "2026-10-21T10:28:44.960Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Focused correction removes the extra cursor increment and updates the comment to match the next-page boundary. All 62 targeted walks, invalid-input and direct-index checks pass without mutating input. Acceptance script and package remain unchanged; unrelated behavior stays intact.", "reviewInputBytes": 4145, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review; full selected text totals4145bytes, below20KiB.", "Only production exports executed:48 varied size/limit walks,6 invalid limits,3 malformed cursors and5 direct-index cases. Walks also verify input nonmutation.", "Acceptance script/package seed equality verified. No official acceptance rerun, candidate edits or dependencies installed. Seed cursor-parser permissiveness was not broadly audited; targeted checks are not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/c379c1f0b5.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/postmortem.md", "nativeTokensPerSec": 7.5, "nativeGeneratedTokens": 2421, "promptTokensPerSec": 497.98, "usageSource": "engine-log", "durationMs": 394931, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.247650146484375 }, "peakRssMb": 7382, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-symptom-debug", "model": "ornith1.5-9b-q4", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T10:29:34.946Z", "finishedAt": "2026-10-21T10:41:08.381Z", "success": false, "failureClass": "model", "reason": "no real progress for 606s — model busy but not delivering (hard digest stuck at 13f02b607717d994)", "modelEligible": true, "fixedComposite": 3.8, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-code-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "No artifact", "justification": "Final pagination module is byte-identical to the buggy seed; no produced correction exists. Four successful file reads are recorded, but no write. Acceptance script and package remain unchanged. The harness stopped a still-generating turn after606seconds without hard progress. Fixture code receives no quality scores.", "reviewInputBytes": 4155, "protocolDeviation": false, "reviewLimitations": [ "No artifact axes or mean assigned. Known fixture source is not model output; complete workspace inventory found no alternative correction.", "Raw recorded tool events, rather than incomplete persisted chat or distilled scenes, establish four successful reads. Native generation continued until expected harness shutdown; the frozen model-stuck classification remains unchanged.", "Model identity was known during review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/d30d12396c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/postmortem.md", "nativeTokensPerSec": 4.75, "nativeGeneratedTokens": 140, "promptTokensPerSec": 559.89, "usageSource": "engine-log", "durationMs": 693435, "toolCalls": 4, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2430877685546875 }, "peakRssMb": 6672, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-symptom-debug", "model": "qwen3.5-4b-q4", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T10:42:00.395Z", "finishedAt": "2026-10-21T10:48:32.108Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, accept-passes)", "modelEligible": true, "fixedComposite": 9.2, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line cursor fix passes all62 targeted pagination, invalid-input and direct-index checks; acceptance script and package remain intact. The adjacent comment still says to resume one past the cursor, contradicting the corrected behavior. Final source provides the requested focused repair.", "reviewInputBytes": 4151, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded complete selected input4151bytes, below20KiB. Production exports only:48 walks,6 invalid limits,3 malformed cursors,5 direct indices.", "No candidate edits or official gate rerun. Script/package seed equality verified. Cursor checks are targeted, not exhaustive.", "Frozen PASS/fixed9.2 is preserved. Terminal sniff reports seed-size1193bytes and lacks paginate-changed while final module is1189bytes and corrected. Source order/timestamps suggest an inconsistent polling snapshot, not a failed final repair; this was not isolated in a reproduction. Process diagnostic evidence records the discrepancy separately." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/a40d87af32.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/postmortem.md", "nativeTokensPerSec": 12.63, "nativeGeneratedTokens": 4250, "promptTokensPerSec": 652.79, "usageSource": "engine-log", "durationMs": 391713, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.24835205078125 }, "peakRssMb": 5007, "peakGpuMemMb": 5720, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-symptom-debug", "model": "underdog-saluki-27b-iq2-mix", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T10:49:32.814Z", "finishedAt": "2026-10-21T10:58:33.375Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Minimal repair removes the cursor offset and clearly documents that the token marks the next item to fetch. All62 varied-walk, invalid-input and direct-index probes pass; input remains unchanged. Acceptance script and package are untouched, preserving the requested scope.", "reviewInputBytes": 4168, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded complete selected input4168bytes, below20KiB.", "Production exports exercised through48 varied walks,6 invalid limits,3 malformed cursors and5 direct-index cases. Walks check input nonmutation.", "Acceptance script/package unchanged. No candidate edits or official acceptance rerun. Targeted checks do not exhaust cursor parsing or arbitrary malformed types." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9535ee2140.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/postmortem.md", "nativeTokensPerSec": 2.89, "nativeGeneratedTokens": 1040, "promptTokensPerSec": 140.59, "usageSource": "engine-log", "durationMs": 540561, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2327117919921875 }, "peakRssMb": 9746, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-data-wrangle", "model": "gemma4-12b-q4", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T10:59:26.205Z", "finishedAt": "2026-10-21T11:11:33.165Z", "success": false, "failureClass": "model", "reason": "no real progress for 604s — model busy but not delivering (hard digest stuck at 059fc5a11b11159c)", "modelEligible": true, "fixedComposite": 3.5, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-data-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "No artifact", "justification": "No customers.json or cleaning script exists in the final workspace; only seeded files remain, with all three raw CSVs unchanged. The output directory was created, but a streamed write hit the8192-token cap and its malformed incomplete arguments were rejected. Partial arguments are not a delivered artifact and receive no quality scores.", "reviewInputBytes": 0, "protocolDeviation": false, "reviewLimitations": [ "No output artifact axes or mean assigned; no artifact text input. Metadata/hashes preserve the seven expected project files and absence of the output.", "Recorded tools confirm three successful CSV reads and directory creation. Native/provider logs record a cap-truncated rejected write and incremental-edit recovery hint before the no-progress stop. Raw partial content and model reasoning were not copied or scored.", "Candidate scripts were not executed, no dependencies installed and no trial retried. Frozen model-stuck classification and performance counters remain unchanged; the final incomplete generation is not fully represented in completed-request token totals.", "Model identity was known during review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/5537df05ad.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/postmortem.md", "nativeTokensPerSec": 14.84, "nativeGeneratedTokens": 8823, "promptTokensPerSec": 396.53, "usageSource": "engine-log", "durationMs": 726960, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.261260986328125 }, "peakRssMb": 13270, "peakGpuMemMb": 8616, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-data-wrangle", "model": "gemma4-e4b-q4", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T11:12:45.478Z", "finishedAt": "2026-10-21T11:49:14.005Z", "success": false, "failureClass": "model", "reason": "retry loop (fast-path): sniff \"data-wrangle:5:targetnone:frcuprhp:rp0:rf0\" stuck for 8m — 3 re-writes (3 tool calls) without sniff movement. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 3.6, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 2, "completeness": 3, "traceability": 1, "outputUsability": 2 }, "advisoryMean": 2, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Valid JSON contains38 of56 customers but replaces every source id with an email, changes Sofia's name, reverses Rosa's date and has four sort inversions. Raw inputs/checker are intact. The selected .mjs helper uses undefined require and references out-of-scope records/file variables; it cannot reproduce a valid transformation as supplied.", "reviewInputBytes": 15965, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. All selected output/helper/grounding text read in full,15965bytes; raw CSVs and checker verified against frozen seeds.", "Independent reference derives56 expected customers from original CSVs, honoring quoted commas, source-specific date formats and newest normalized-email records. Final output has38 unique customers,18 missing,38 non-source ids,two additional field mismatches and four adjacent sort inversions.", "Only the latest-written selected helper was inspected; two alternative helpers were intentionally not read or scored. Their filenames/lengths and tool-call metadata were inventoried, not their contents.", "Candidate scripts were not executed or repaired. require in .mjs, block-local records returned outside scope, and file used outside its loop are source findings. Official failure/gates/fixed score unchanged; no data file modified." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/38b493a33f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/postmortem.md", "nativeTokensPerSec": 21.15, "nativeGeneratedTokens": 37406, "promptTokensPerSec": 641.15, "usageSource": "engine-log", "durationMs": 2188527, "toolCalls": 20, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.320404052734375 }, "peakRssMb": 6533, "peakGpuMemMb": 4058, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-data-wrangle", "model": "qwen3.5-9b-q4", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T11:50:13.633Z", "finishedAt": "2026-10-21T12:11:50.569Z", "success": false, "failureClass": "model", "reason": "retry loop (fast-path): sniff \"data-wrangle:4:targetnone:frf6ms86:rp0:rf0\" stuck for 10m — 3 re-writes (7 tool calls) without sniff movement. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 3.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 5, "traceability": 3, "outputUsability": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All56 customers appear in58 rows, but Zoe and Omar remain duplicated. Alice has the wrong retained id/date, and an Omar row combines an old id with the newer date. Six sort inversions remain. The helper pads days with '10' and keeps original email case in output, so it does not reliably reproduce the requested normalization. Seeds are intact.", "reviewInputBytes": 15246, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full selected text15246bytes, below20KiB. Raw CSVs and checker are unchanged.", "Independent source-derived comparison finds56 unique expected emails,58 total rows,two duplicate emails,five field mismatches against retained records and six adjacent sort inversions. No invented source ids, but some existing ids are paired with incorrect retained dates.", "The helper was inspected, not executed. Day padding with the string10 and failure to assign normalizedEmail to output are explicit source defects; no repaired variant or hypothetical generated file was scored.", "JSON validity and exact output fields were checked independently; frozen gate/fixed score and all candidate files remain unchanged. Candidate runtime and every generic CSV edge case were not exhaustively tested." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/00468ab8be.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/postmortem.md", "nativeTokensPerSec": 10.65, "nativeGeneratedTokens": 12523, "promptTokensPerSec": 374.34, "usageSource": "engine-log", "durationMs": 1296936, "toolCalls": 16, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2485809326171875 }, "peakRssMb": 8323, "peakGpuMemMb": 8100, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-data-wrangle", "model": "ornith1.5-9b-q4", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T12:12:54.162Z", "finishedAt": "2026-10-21T12:24:32.473Z", "success": false, "failureClass": "model", "reason": "no real progress for 605s — model busy but not delivering (hard digest stuck at 139a4c9ae32c7d7d)", "modelEligible": true, "fixedComposite": 3.5, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-data-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "No artifact", "justification": "No customers.json or cleaning script was produced. The final project contains only seven seeded files; raw CSVs are unchanged. Four successful reads retrieved the inputs and checker, then generation continued without a write until the605-second no-progress stop. No artifact quality scores assigned.", "reviewInputBytes": 0, "protocolDeviation": false, "reviewLimitations": [ "No advisory axes or mean: an unchanged seed and incomplete generation are not produced artifacts.", "Model identity known. Process diagnostics include successful read metadata and native generation/shutdown lines, not raw model reasoning or hypothetical output.", "Expected shutdown is not GPU OOM. Frozen model-stuck outcome and completed-request performance counters are unchanged.", "Model identity was known during review." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/b64dc50f6c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/postmortem.md", "nativeTokensPerSec": 4.75, "nativeGeneratedTokens": 149, "promptTokensPerSec": 560.13, "usageSource": "engine-log", "durationMs": 698311, "toolCalls": 4, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2440338134765625 }, "peakRssMb": 6653, "peakGpuMemMb": 7254, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-data-wrangle", "model": "qwen3.5-4b-q4", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T12:25:26.300Z", "finishedAt": "2026-10-21T13:20:45.672Z", "success": false, "failureClass": "model", "reason": "retry loop (fast-path): sniff \"data-wrangle:4:targetnone:frszsiwu:rp0:rf0\" stuck for 16m — 3 re-writes (14 tool calls) without sniff movement. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 2.4, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 4, "traceability": 3, "outputUsability": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Valid JSON has 57 rows for 55 of 56 customers: Viktor is missing, Omar/Robert are duplicated with stale values, and 17 adjacent pairs are unsorted. The latest cleaner fails syntax checking on duplicate parseCSV declarations. The overwritten checker uses require in .mjs and drops source-membership and expected-count checks. Raw inputs remain intact.", "reviewInputBytes": 18603, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. All selected source and grounding text read in full within the combined20KiB cap. Raw CSVs match frozen seeds; checker was overwritten by the candidate.", "Independent reference finds57 rows,55 unique expected customers,missing Viktor,two duplicate emails,five retained-record field mismatches and17 adjacent sort inversions. No invented source ids; stale Omar additionally mis-converts the original European date.", "Only the latest cleaner and overwritten checker were inspected. Alternative helpers and unrelated shared weekly digest were excluded before reading; their contents are not scored or characterized.", "node --check on the unchanged cleaner reports duplicate parseCSV declaration at line24. No candidate code was executed or repaired. Wrong regex match destructuring, skipped legacy banner, extra dateStr and unnormalized stored email are source findings beyond the blocking syntax error, not claims about an executed repaired program.", "Rewritten checker uses require in .mjs, replaces exact source-id membership with an A-L regex and omits expected row count. Original checker is retained as counted grounding. Official failure, fixed score and all candidate artifacts remain unchanged. No broad CSV edge-case execution or official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/9b0de6ab8a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/postmortem.md", "nativeTokensPerSec": 16.34, "nativeGeneratedTokens": 33585, "promptTokensPerSec": 424.39, "usageSource": "engine-log", "durationMs": 3319372, "toolCalls": 45, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.527984619140625 }, "peakRssMb": 11536, "peakGpuMemMb": 5722, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-data-wrangle", "model": "underdog-saluki-27b-iq2-mix", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T13:21:52.575Z", "finishedAt": "2026-10-21T14:04:37.132Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on out/customers.json; missing signals stayed unchanged: row-count. Last failure: expected exactly 56 records after dedup, got 55", "modelEligible": true, "fixedComposite": 5.1, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 4, "completeness": 4, "traceability": 5, "outputUsability": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The JSON preserves every field of 55 unique customers but omits Sofia and has eight sorting inversions. Raw inputs/checker are unchanged. Its readable cleaner produces the exact56-row reference in an isolated unchanged-code replay, exposing a delivery/execution gap. That diagnostic success does not replace the incomplete final file.", "reviewInputBytes": 13398, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Selected artifact and grounding text13398bytes, within20KiB. Raw inputs/checker match frozen seeds;55 retained records have no field mismatches, no duplicates or invented ids. Missing Sofia and eight adjacent sort inversions remain.", "Initial source-only evidence was saved before a separate exact-source replay: advisory-data-evidence/87f441de06-helper-replay.json. The replay copies only the inspected helper and exact raw CSVs into a fresh directory with an existing out folder, matching the final project. No candidate source edits or added dependencies.", "Replay output exactly matches an independently derived56-row reference, but is diagnostic only. Its generated JSON was not substituted for or scored as the final artifact; only comparison results/hash were inspected. Supplied CSVs only, no broad escaped-quote/malformed-input testing.", "The preserved trial workspace, official failure/gates/fixed score and performance remain unchanged. No official acceptance rerun. Actual model-facing tool surface omitted run_nodejs_script; two incorrect installed-script calls were rejected. No permission expansion or model tuning was performed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/87f441de06.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/postmortem.md", "nativeTokensPerSec": 9.72, "nativeGeneratedTokens": 22564, "promptTokensPerSec": 173.15, "usageSource": "engine-log", "durationMs": 2564556, "toolCalls": 21, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.233642578125 }, "peakRssMb": 13664, "peakGpuMemMb": 10626, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-data-wrangle", "model": "qwen3.8-27b-q2", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 3, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T14:05:46.456Z", "finishedAt": "2026-10-21T14:45:19.825Z", "success": false, "failureClass": "model", "reason": "retry loop (long-path): sniff \"data-wrangle:7:targetnone:fr1rezbry:rp0:rf0\" stuck for 12m — team idle (2 new tool calls) with no sniff movement for 12m. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 5.8, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 4, "completeness": 5, "traceability": 4, "outputUsability": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All56 unique customers and required fields are present, with intact source ids/names and no duplicates. Alice's retained date is reversed and six adjacent pairs are unsorted. The latest cleaner has explicit source-date handling and sorting, but the delivered JSON does not match that logic. Seeds/checker are unchanged; helper execution was not tested.", "reviewInputBytes": 15067, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Complete selected artifact and grounding text15067bytes within20KiB. Independent reference derives56 expected records from frozen raw CSVs; all seeds/checker unchanged.", "The older tools/clean_customers.mjs contents were excluded before reading to respect the combined cap. Only the latest helper is assessed. No candidate script execution, source repair, replay or official gate rerun.", "Six adjacent sort inversions and one date mismatch are measured in the delivered JSON. Static source inspection suggests the latest helper would normalize/sort differently; this is not a verified execution result.", "Three expected memory-pressure native exits occurred after foreground work became idle; preserve the runtime caveat separately from artifact quality and the frozen model-stuck classification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/4c44488e89.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/postmortem.md", "nativeTokensPerSec": 6.61, "nativeGeneratedTokens": 12759, "promptTokensPerSec": 467.06, "usageSource": "engine-log", "durationMs": 2373369, "toolCalls": 11, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.283843994140625 }, "peakRssMb": 15416, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-incident-postmortem", "model": "gemma4-e4b-q4", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T14:46:31.816Z", "finishedAt": "2026-10-21T14:48:17.744Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 3, "clarity": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct core cause and required sections, but recovery is placed a minute early and exact peaks are overstated. Broad gateway/pipeline conclusions exceed the evidence. Three actions have TBD deadlines and lack verification criteria; a joined heading and repetitive deep dive remain.", "reviewInputBytes": 13406, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/f6693144dc.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/postmortem.md", "nativeTokensPerSec": 165.81, "nativeGeneratedTokens": 9976, "promptTokensPerSec": 4331.29, "usageSource": "engine-log", "durationMs": 105928, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.335601806640625 }, "peakRssMb": 6123, "peakGpuMemMb": 4143, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-incident-postmortem", "model": "qwen3.5-9b-q4", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T14:49:12.029Z", "finishedAt": "2026-10-21T14:52:29.534Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 2, "functionalCompleteness": 5, "actionability": 5, "clarity": 3 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Useful owned actions and correct central mechanism, but several metric/time claims conflict with the sources:14:54 values, saturation window and below1% recovery. It calls14:46–14:53 a23-minute deployment and asserts missing controls without evidence. Repetition and inconsistent chronology undermine an otherwise complete report.", "reviewInputBytes": 16492, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fdfc6d5b21.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/postmortem.md", "nativeTokensPerSec": 80.05, "nativeGeneratedTokens": 9119, "promptTokensPerSec": 2388.23, "usageSource": "engine-log", "durationMs": 197505, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2992401123046875 }, "peakRssMb": 10540, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-incident-postmortem", "model": "ornith1.5-9b-q4", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T14:53:27.431Z", "finishedAt": "2026-10-21T14:59:50.989Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 6, "actionability": 4, "clarity": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Explicit unknown impacts and hotfix-source reconciliation are useful. However, saturation/request-rate windows and monitoring chronology are wrong; absent rollback, approval and canary controls are asserted without evidence. Owned actions and deadlines are present, but some citations do not support their claims and completion criteria are limited.", "reviewInputBytes": 16034, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/1a025ebfa0.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/postmortem.md", "nativeTokensPerSec": 54.02, "nativeGeneratedTokens": 16201, "promptTokensPerSec": 2549.25, "usageSource": "engine-log", "durationMs": 383558, "toolCalls": 16, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2877960205078125 }, "peakRssMb": 9652, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-incident-postmortem", "model": "qwen3.5-4b-q4", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T15:01:00.300Z", "finishedAt": "2026-10-21T15:02:52.706Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 2, "functionalCompleteness": 4, "actionability": 4, "clarity": 4 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core cause and sections are readable, with owned dated actions. It invents no data loss/security breach and eventual completion of all payments, mixes recovery timestamps and misattributes hotfix start. Self-merge bypass and absent validation are unproven. These impact claims materially undermine the requested evidence-based report.", "reviewInputBytes": 13684, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d0a4c22299.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/postmortem.md", "nativeTokensPerSec": 113.3, "nativeGeneratedTokens": 6717, "promptTokensPerSec": 3504.59, "usageSource": "engine-log", "durationMs": 112406, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2901458740234375 }, "peakRssMb": 8785, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem", "model": "underdog-saluki-27b-iq2-mix", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T15:03:57.776Z", "finishedAt": "2026-10-21T15:14:44.053Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 6, "actionability": 5, "clarity": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct21-minute recovery, explicit unknown impacts and reconciled hotfix timestamps provide a useful evidence chain. It extends99% saturation through14:52 incorrectly and calls14:34:11 within one minute of14:32 completion. Owned dated actions are practical but imply unproven missing controls, and the merge-to-rollout gap is not shown causal.", "reviewInputBytes": 16181, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/79046b4f8e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/postmortem.md", "nativeTokensPerSec": 12.46, "nativeGeneratedTokens": 7190, "promptTokensPerSec": 659.57, "usageSource": "engine-log", "durationMs": 646277, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3229217529296875 }, "peakRssMb": 9761, "peakGpuMemMb": 10709, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-incident-postmortem", "model": "qwen3.8-27b-q2", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T15:15:38.088Z", "finishedAt": "2026-10-21T16:15:15.665Z", "success": false, "failureClass": "model", "reason": "chat stalled for 305s — daemon reachable but issuing no model turns (soft digest stuck at 28ee7483ee97a6be) (re-engage nudge was sent and ignored)", "modelEligible": true, "fixedComposite": 5.9, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core metrics, unknown impacts and hotfix-source reconciliation are useful. The impact window ends a minute early, monitoring is placed before14:58, and missing alerts/revalidation are inferred without evidence. Nine summary sentences and relative deadlines need refinement. Clear sections and owned actions remain useful despite the failed gate.", "reviewInputBytes": 17235, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/45668bd566.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/postmortem.md", "nativeTokensPerSec": 5.82, "nativeGeneratedTokens": 18065, "promptTokensPerSec": 461.06, "usageSource": "engine-log", "durationMs": 3577577, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.323974609375 }, "peakRssMb": 15867, "peakGpuMemMb": 11689, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-incident-postmortem", "model": "gemma4-12b-q4", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T16:16:26.048Z", "finishedAt": "2026-10-21T16:18:41.149Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Captures the timeout cause, peaks, required sections and owned actions. Summary mitigation at14:53 conflicts with later recovery; source chronology is not reconciled. It calls the author the reviewer and overstates immediate504 behavior/review bypass. Unexplained October/November deadlines and an appended repetitive deep dive weaken usefulness.", "reviewInputBytes": 13679, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The complete report and five unchanged grounding files were read within a combined 20 KiB limit. Only the report is scored as model output.", "Prose and source fidelity were inspected directly; no rendered-document or exhaustive citation audit.", "Source chronology discrepancies are distinguished from unsupported claims and arithmetic errors." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fc6bcce7e8.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/postmortem.md", "nativeTokensPerSec": 118.74, "nativeGeneratedTokens": 7918, "promptTokensPerSec": 1854.1, "usageSource": "engine-log", "durationMs": 135101, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3682403564453125 }, "peakRssMb": 16504, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-ops-runbook-anomaly", "model": "qwen3.5-9b-q4", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T16:19:44.060Z", "finishedAt": "2026-10-21T16:20:51.616Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact values are recorded under STEP1–4, and the concise halt report contrasts fresh/under24 with stale/71. Source reads and log writes preserve the step sequence; later-step files are absent. A final log rewrite follows the halt report and an instruction reread. A complete, usable record without exceptional polish.", "reviewInputBytes": 4831, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review of both complete final reports, all seven unchanged seed files, and compact tool-event metadata. Every selected text input, including metadata, counts toward the combined20KiB cap.", "Tool operation order was read directly from the recording; no earlier write contents or raw model reasoning were included. Earlier log contents are not reconstructed, but no write occurred between the identified later reads.", "Workspace inventory confirms no maintenance/step5.done or decommission-list.txt. The frozen gate and score are preserved; no checker rerun or candidate edits. Source files and trace are grounding, not produced reports." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d5145b9053.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/postmortem.md", "nativeTokensPerSec": 80.29, "nativeGeneratedTokens": 1939, "promptTokensPerSec": 2399.09, "usageSource": "engine-log", "durationMs": 67556, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2756500244140625 }, "peakRssMb": 9804, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-ops-runbook-anomaly", "model": "ornith1.5-9b-q4", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T16:21:45.687Z", "finishedAt": "2026-10-21T16:22:59.407Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 6 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct values, explicit halt, source references and a clear expected-versus-observed table. No later-step files exist. The final log is complete, but no log update separates the checksum, config and backup checks, contrary to the after-each-step rule. Strong presentation does not erase that procedural gap.", "reviewInputBytes": 5049, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review of both complete final reports, all seven unchanged seed files, and compact tool-event metadata. Every selected text input, including metadata, counts toward the combined20KiB cap.", "Tool operation order was read directly from the recording; no earlier write contents or raw model reasoning were included. Earlier log contents are not reconstructed, but no write occurred between the identified later reads.", "Workspace inventory confirms no maintenance/step5.done or decommission-list.txt. The frozen gate and score are preserved; no checker rerun or candidate edits. Source files and trace are grounding, not produced reports." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c9a224af82.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/postmortem.md", "nativeTokensPerSec": 56.48, "nativeGeneratedTokens": 2606, "promptTokensPerSec": 2627.47, "usageSource": "engine-log", "durationMs": 73720, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.331298828125 }, "peakRssMb": 6779, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-ops-runbook-anomaly", "model": "qwen3.5-4b-q4", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T16:23:53.995Z", "finishedAt": "2026-10-21T16:24:44.025Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Final STEP1–4 records and halt values are correct, with clear instructions not to proceed and no downstream files. However, checksum, config and backup were read without intervening log writes; the final rewrite followed the halt report and a reread. The useful final artifact does not establish the required incremental audit trail.", "reviewInputBytes": 4652, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review of both complete final reports, all seven unchanged seed files, and compact tool-event metadata. Every selected text input, including metadata, counts toward the combined20KiB cap.", "Tool operation order was read directly from the recording; no earlier write contents or raw model reasoning were included. Earlier log contents are not reconstructed, but no write occurred between the identified later reads.", "Workspace inventory confirms no maintenance/step5.done or decommission-list.txt. The frozen gate and score are preserved; no checker rerun or candidate edits. Source files and trace are grounding, not produced reports." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fbc55aaa60.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/postmortem.md", "nativeTokensPerSec": 112.44, "nativeGeneratedTokens": 1878, "promptTokensPerSec": 3617.38, "usageSource": "engine-log", "durationMs": 50030, "toolCalls": 11, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2650604248046875 }, "peakRssMb": 6820, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly", "model": "underdog-saluki-27b-iq2-mix", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T16:25:40.747Z", "finishedAt": "2026-10-21T16:30:02.439Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact ordered STEP1–4 values and source paths, with a clear fresh/under24 versus stale/71 comparison. The halt report names both failed conditions and explicitly skips steps5–6; their files are absent. Trace shows log writes between successful checks, then a final rewrite after the halt and reread. Complete and usable.", "reviewInputBytes": 4966, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct review. Both final reports, seven seed files and compact tool metadata are counted under the combined20KiB limit.", "Tool operation order and final files were inspected; earlier write contents and model reasoning were not read. All seeds unchanged, prohibited downstream outputs absent.", "No official checker rerun, source edits or changed gate outcome. Validation was a recorded candidate action, not a newly executed review test." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fc1a0352bd.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/postmortem.md", "nativeTokensPerSec": 12.33, "nativeGeneratedTokens": 2093, "promptTokensPerSec": 618.19, "usageSource": "engine-log", "durationMs": 261692, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.26007080078125 }, "peakRssMb": 14654, "peakGpuMemMb": 10711, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-ops-runbook-anomaly", "model": "qwen3.8-27b-q2", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T16:31:00.676Z", "finishedAt": "2026-10-21T16:37:24.564Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact STEP1–4 values, a clear expected-versus-observed halt report, and explicit unexecuted downstream steps. Their files are absent. However, all source checks were read before the first runlog write, so the requested after-each-step recording was not followed. The final record is concise and useful.", "reviewInputBytes": 3680, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final reports, seven unchanged seeds and compact tool metadata all count toward the20KiB cap.", "Actual operation order confirms all state reads preceded the first runlog write. Earlier write contents and reasoning were not reviewed. No downstream markers exist.", "Frozen passing gate and composite10 remain unchanged. No checker rerun, candidate edits or model/runtime change." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/33341c664d.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/postmortem.md", "nativeTokensPerSec": 5.01, "nativeGeneratedTokens": 1450, "promptTokensPerSec": 459.96, "usageSource": "engine-log", "durationMs": 383888, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2584991455078125 }, "peakRssMb": 12544, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-ops-runbook-anomaly", "model": "gemma4-12b-q4", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T16:38:51.706Z", "finishedAt": "2026-10-21T16:40:17.915Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact STEP1–4 values and clear fresh/under24 versus stale/71 halt details. Log writes separate the successful checks, and no downstream files exist. Step4 is added during final repair. Concise, usable final records; the earlier writes were not read, so their intermediate contents are not certified.", "reviewInputBytes": 4447, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final reports, seven unchanged seeds and compact actual tool metadata are all counted under20KiB.", "Tool timestamps establish operation order; earlier write contents and reasoning were not reviewed. No later-step sentinel files exist.", "Frozen deterministic outcomes and performance remain unchanged. No checker rerun or candidate edit." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4a68e12831.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/postmortem.md", "nativeTokensPerSec": 121.02, "nativeGeneratedTokens": 1481, "promptTokensPerSec": 1764.59, "usageSource": "engine-log", "durationMs": 86209, "toolCalls": 18, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3005523681640625 }, "peakRssMb": 17090, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-ops-runbook-anomaly", "model": "gemma4-e4b-q4", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T16:41:29.701Z", "finishedAt": "2026-10-21T16:42:48.070Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact ordered values and a grounded halt report with explicit failed conditions and stopped execution. Writes separate the successful checks; downstream files are absent. The final log repair restores all four STEP headings. Minimal and usable, with no exceptional polish.", "reviewInputBytes": 5014, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final reports, seven unchanged seeds and compact actual tool metadata are all counted under20KiB.", "Tool timestamps establish operation order; earlier write contents and reasoning were not reviewed. No later-step sentinel files exist.", "Frozen deterministic outcomes and performance remain unchanged. No checker rerun or candidate edit." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/25d6692d2c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/postmortem.md", "nativeTokensPerSec": 193.51, "nativeGeneratedTokens": 5496, "promptTokensPerSec": 4490.06, "usageSource": "engine-log", "durationMs": 78369, "toolCalls": 21, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3035888671875 }, "peakRssMb": 6300, "peakGpuMemMb": 4143, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-plan-and-estimate", "model": "ornith1.5-9b-q4", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T16:43:42.972Z", "finishedAt": "2026-10-21T16:44:40.456Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear internal ownership, sign-off before mover booking, and useful network checks. However, no task executes the physical move or completes the workshop build; the workshop task stops at scope approval. Network acceptance is described on move day despite the pre-day-one objective. Readable structure leaves material execution work unassigned.", "reviewInputBytes": 3795, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/40a973326b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/postmortem.md", "nativeTokensPerSec": 58.1, "nativeGeneratedTokens": 2047, "promptTokensPerSec": 2803.54, "usageSource": "engine-log", "durationMs": 57484, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.24505615234375 }, "peakRssMb": 6657, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-plan-and-estimate", "model": "qwen3.5-4b-q4", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T16:45:34.574Z", "finishedAt": "2026-10-21T16:47:20.508Z", "success": false, "failureClass": "model", "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 2 while the failing detail kept changing; latest checked file plan.md. Last failure: plan table: the plan table is missing column(s): task, owner, depends on, done when (found: ID | Assumption)", "modelEligible": true, "fixedComposite": 3.8, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 3, "clarity": 4 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The work table has the requested columns, but assigns T9 to forbidden contractor Ola and T11 to multiple owners. Final move readiness omits packing/workshop dependencies and says the destination studio is empty. The evaluator selects the earlier Assumptions table; a diagnostic selection probe still fails on Ola. Official failure is preserved.", "reviewInputBytes": 4326, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/56914e9354.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/postmortem.md", "nativeTokensPerSec": 122.23, "nativeGeneratedTokens": 5531, "promptTokensPerSec": 3780.42, "usageSource": "engine-log", "durationMs": 105934, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2862091064453125 }, "peakRssMb": 9368, "peakGpuMemMb": 5807, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate", "model": "underdog-saluki-27b-iq2-mix", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T16:48:14.547Z", "finishedAt": "2026-10-21T16:50:57.993Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct internal owners, signed layout before booking, workshop acceptance and network tests. The plan omits an explicit physical move execution task, and network installation does not depend on its hardware-order task. Final readiness therefore does not establish that all moving work occurred. Concise and useful, with an incomplete execution chain.", "reviewInputBytes": 4251, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4d16f0ba9a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/postmortem.md", "nativeTokensPerSec": 13.22, "nativeGeneratedTokens": 1511, "promptTokensPerSec": 698.55, "usageSource": "engine-log", "durationMs": 163446, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.23468017578125 }, "peakRssMb": 9723, "peakGpuMemMb": 10711, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-plan-and-estimate", "model": "qwen3.8-27b-q2", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T16:51:52.693Z", "finishedAt": "2026-10-21T16:59:38.562Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Covers actual moving, an internally owned workshop build, formal approval before booking, and dated pre-day-one network tests. Dependencies need tightening: the workshop can start before sign-off, while the risk instruction to lock ISP cutover before booking is absent from that dependency chain. Practical and readable overall.", "reviewInputBytes": 5674, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5e1b2936ee.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/postmortem.md", "nativeTokensPerSec": 3.94, "nativeGeneratedTokens": 1566, "promptTokensPerSec": 500.17, "usageSource": "engine-log", "durationMs": 465869, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.235748291015625 }, "peakRssMb": 11920, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-plan-and-estimate", "model": "gemma4-12b-q4", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T17:00:32.492Z", "finishedAt": "2026-10-21T17:01:08.437Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct roster ownership and approval-before-booking dependency, with physical move and network installation tasks. Workshop work stops at confirming a schedule, without build completion or acceptance. No final readiness check joins the workstreams, and the pre-day-one network deadline is not enforced by the task chain. Clear but incomplete execution coverage.", "reviewInputBytes": 2890, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/8357a6a519.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/postmortem.md", "nativeTokensPerSec": 105.95, "nativeGeneratedTokens": 1021, "promptTokensPerSec": 1894.08, "usageSource": "engine-log", "durationMs": 35945, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.262298583984375 }, "peakRssMb": 10669, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-plan-and-estimate", "model": "gemma4-e4b-q4", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T17:02:02.685Z", "finishedAt": "2026-10-21T17:02:30.063Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 4, "actionability": 3, "clarity": 5 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The signed-layout booking dependency and physical move are clear. It adds an unsupported Kolbrun approval, repeatedly mismatches IT/design/operations roles, and leaves workshop execution and print/scan installation unassigned. Network and move tasks exist, but the all-areas objective exceeds the actual delivery chain. Readable headings and table.", "reviewInputBytes": 3511, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/381910c703.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/postmortem.md", "nativeTokensPerSec": 175.88, "nativeGeneratedTokens": 1807, "promptTokensPerSec": 4707.42, "usageSource": "engine-log", "durationMs": 27378, "toolCalls": 1, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.24981689453125 }, "peakRssMb": 4467, "peakGpuMemMb": 4143, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-plan-and-estimate", "model": "qwen3.5-9b-q4", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T17:03:23.826Z", "finishedAt": "2026-10-21T17:04:09.534Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Covers layout sign-off, booking, physical move, internally owned workshop completion, and network/print tests with observable criteria. Dates and performance targets are planning choices, not source facts. Final orientation omits the workshop dependency and some network installation steps remain implicit, but the overall plan is usable and meets the brief.", "reviewInputBytes": 3817, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all two brief/team files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/1d4a5b87af.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/postmortem.md", "nativeTokensPerSec": 54.16, "nativeGeneratedTokens": 1054, "promptTokensPerSec": 2556.59, "usageSource": "engine-log", "durationMs": 45708, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2496337890625 }, "peakRssMb": 7357, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-4b-q4-core-conflict-synthesis", "model": "qwen3.5-4b-q4", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T17:05:03.062Z", "finishedAt": "2026-10-21T17:06:32.679Z", "success": false, "failureClass": "model", "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 date) is not source-bound: re-read memo-product.md and memo-engineering.md, then state each file's observed launch date field in one Conflicts entry and explain which source controls it - conflict (launch budget) is not source-bound: re-read memo-product.md and finance.csv, then state each file's observed launch budget field in one Conflicts entry and explain which source controls it - conflict (launch DRI) is not source-bound: re-read memo-oldplan.md and org.md, then state each file's observed launch DRI field in one Conflicts entry and explain which source controls it - only 1 substantively cited claim(s) — name the seeded source file as a backticked filename inside the ", "modelEligible": true, "fixedComposite": 4.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winning date, budget allocation and owners; all three historical values stay in the conflict table and source priority is explained. Numbered references resolve through a Sources list, but do not satisfy the explicitly requested backticked filenames for six claims. The DRI row's Product heading is misleading. Useful synthesis with incomplete contract compliance.", "reviewInputBytes": 2635, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all six seed files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7babcba5c5.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/postmortem.md", "nativeTokensPerSec": 118.19, "nativeGeneratedTokens": 4807, "promptTokensPerSec": 3825.93, "usageSource": "engine-log", "durationMs": 89617, "toolCalls": 30, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.342559814453125 }, "peakRssMb": 8029, "peakGpuMemMb": 5809, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis", "model": "underdog-saluki-27b-iq2-mix", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T17:07:26.612Z", "finishedAt": "2026-10-21T17:18:34.766Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winners, complete budget allocation, explicit source-bound reconciliation and quarantined historical values. Current ownership and unchanged feature scope are clear. Open questions sensibly separate unknown meeting cadence and delivery detail, though the budget question presumes a line-item reduction. Concise, usable synthesis meeting the brief.", "reviewInputBytes": 3762, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review. Full final document and all six seed files count toward the combined20KiB cap. Only the authored document is scored.", "Direct prose, source fidelity and dependency/claim review; no rendered-document inspection or operational execution of the plan.", "All official gates, scores and classifications remain unchanged. Diagnostic probes use only already-reviewed text in memory; no candidate or active runtime edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/3c4a7c3b12.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/postmortem.md", "nativeTokensPerSec": 12.82, "nativeGeneratedTokens": 5449, "promptTokensPerSec": 678.37, "usageSource": "engine-log", "durationMs": 668154, "toolCalls": 24, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3021697998046875 }, "peakRssMb": 17522, "peakGpuMemMb": 10713, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.8-27b-q2-core-conflict-synthesis", "model": "qwen3.8-27b-q2", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T17:19:42.363Z", "finishedAt": "2026-10-21T17:32:46.449Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winners, allocation and source-bound conflicts. It asserts historical Tuesday meetings are still in force without current support, and treats one contractor budget line as possible evidence that a contractor was dropped. The migration-start question is grounded. Useful synthesis, but adopted assumptions exceed the supplied evidence.", "reviewInputBytes": 3800, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct document review. Complete final synthesis and all six source/style files counted under the combined20KiB cap; only authored synthesis scored.", "Source fidelity, contract compliance and uncertainty handling inspected. No rendered-document review, candidate edits, gate changes or reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c8b2a443cb.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/postmortem.md", "nativeTokensPerSec": 5.92, "nativeGeneratedTokens": 3922, "promptTokensPerSec": 481.37, "usageSource": "engine-log", "durationMs": 784086, "toolCalls": 15, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2833099365234375 }, "peakRssMb": 16794, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-12b-q4-core-conflict-synthesis", "model": "gemma4-12b-q4", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T17:33:49.267Z", "finishedAt": "2026-10-21T17:36:30.349Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Concise correct date, budget, ownership, scope and migration facts, with source filenames on claims and all three explicit resolutions. Historical values stay quarantined. Budget line-item detail is omitted, but the short brief and practical questions meet the requested synthesis without unsupported assertions or exceptional polish.", "reviewInputBytes": 2206, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct document review. Complete final synthesis and all six source/style files counted under the combined20KiB cap; only authored synthesis scored.", "Source fidelity, contract compliance and uncertainty handling inspected. No rendered-document review, candidate edits, gate changes or reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/a9a1313f0e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/postmortem.md", "nativeTokensPerSec": 128.88, "nativeGeneratedTokens": 4475, "promptTokensPerSec": 1958.18, "usageSource": "engine-log", "durationMs": 161082, "toolCalls": 22, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.33184814453125 }, "peakRssMb": 16946, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-gemma4-e4b-q4-core-conflict-synthesis", "model": "gemma4-e4b-q4", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T17:37:34.143Z", "finishedAt": "2026-10-21T17:39:25.540Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 7.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 4, "actionability": 3, "clarity": 4 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winners, allocation and source-priority resolutions with usable citations. Extra Method/Findings structure adds noise; campaign and engineering leads are omitted. The sole Open Questions bullet incorrectly says no other historical context exists, although the old plan specifies Tuesday syncs. It supplies no concrete question to resolve that uncertainty.", "reviewInputBytes": 2912, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct document review. Complete final synthesis and all six source/style files counted under the combined20KiB cap; only authored synthesis scored.", "Source fidelity, contract compliance and uncertainty handling inspected. No rendered-document review, candidate edits, gate changes or reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5ac04a5fc1.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/postmortem.md", "nativeTokensPerSec": 199.85, "nativeGeneratedTokens": 9431, "promptTokensPerSec": 4645.98, "usageSource": "engine-log", "durationMs": 111397, "toolCalls": 22, "redFlags": [ { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/sessions/tamsin", "tool": "(no write_file this turn)", "argsSummary": "content_chars=2619 h2_count=4 preview=\"# Skylark Launch Synthesis — analysis ## Summary This brief consolidates the product, engineering, finance, and organiz…\"", "atTurn": 1, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." } ], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3189239501953125 }, "peakRssMb": 7188, "peakGpuMemMb": 4143, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-qwen3.5-9b-q4-core-conflict-synthesis", "model": "qwen3.5-9b-q4", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T17:40:18.777Z", "finishedAt": "2026-10-21T17:42:45.860Z", "success": false, "failureClass": "model", "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 5 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", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 4, "actionability": 4, "clarity": 4 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct canonical values, allocation and three reconciliations. Most prose uses numbered references despite the explicit six-claim backticked-filename requirement. It labels the old Tuesday cadence N/A without a current replacement and repeats source lists. Useful open questions, but citation compliance and uncertainty handling remain incomplete.", "reviewInputBytes": 3265, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct document review. Complete final synthesis and all six source/style files counted under the combined20KiB cap; only authored synthesis scored.", "Source fidelity, contract compliance and uncertainty handling inspected. No rendered-document review, candidate edits, gate changes or reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/ab3f18083d.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/postmortem.md", "nativeTokensPerSec": 81.78, "nativeGeneratedTokens": 4671, "promptTokensPerSec": 2575.54, "usageSource": "engine-log", "durationMs": 147083, "toolCalls": 32, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3338165283203125 }, "peakRssMb": 14173, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r2-ornith1.5-9b-q4-core-conflict-synthesis", "model": "ornith1.5-9b-q4", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T17:43:39.101Z", "finishedAt": "2026-10-21T17:50:40.639Z", "success": false, "failureClass": "model", "reason": "repair-aborted: tamsin/ab53d5a6 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.", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 4, "clarity": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core winners and conflict resolutions are correct, but the superseded240,000 budget reappears in Open questions. Tuesday meetings are adopted in the plan while another section says they are unconfirmed. Contractor-count speculation, duplicated non-conflicts and repetitive prose weaken an otherwise useful brief. The explicit quarantine contract is not met.", "reviewInputBytes": 4708, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct document review. Complete final synthesis and all six source/style files counted under the combined20KiB cap; only authored synthesis scored.", "Source fidelity, contract compliance and uncertainty handling inspected. No rendered-document review, candidate edits, gate changes or reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/ae6e00bb64.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/postmortem.md", "nativeTokensPerSec": 57.26, "nativeGeneratedTokens": 18196, "promptTokensPerSec": 2652.14, "usageSource": "engine-log", "durationMs": 421538, "toolCalls": 53, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.435394287109375 }, "peakRssMb": 14955, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-tictactoe", "model": "qwen3.5-9b-q4", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T17:51:35.007Z", "finishedAt": "2026-10-21T17:54:56.982Z", "success": false, "failureClass": "model", "reason": "repair-aborted: hamish/5ea14afa 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.", "modelEligible": true, "fixedComposite": 5.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 1, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clicks and keyboard activation throw because the listener passes an Event to a function expecting an element; no move is playable. The tiny pale board sits left of the centered heading/reset, and viewport metadata is absent. Win/draw logic exists but is unreachable. Reset cannot rescue the broken primary interaction.", "reviewInputBytes": 10188, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/92b0724d3a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/postmortem.md", "nativeTokensPerSec": 92.61, "nativeGeneratedTokens": 10642, "promptTokensPerSec": 2517.59, "usageSource": "engine-log", "durationMs": 201975, "toolCalls": 19, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2940521240234375 }, "peakRssMb": 11872, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-tictactoe", "model": "ornith1.5-9b-q4", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T17:56:17.280Z", "finishedAt": "2026-10-21T17:59:19.130Z", "success": true, "failureClass": "pass", "reason": "this-work-e/workspace/index.html passed sniff (signals: name, grid, click, win-detect, js-parses, js-size-ok) + runtime (2 assertion(s))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 4, "codeQuality": 4, "polish": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Coherent dark layout, colored marks and winning-line highlight. Turns, win, draw, duplicate-click guard and terminal lock work, but New game clears marks while leaving old win classes. The next empty board and later draw retain a false highlighted winner. Keyboard buttons and narrow layout work; reset-state cleanup is incomplete.", "reviewInputBytes": 5147, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8c44bfb429.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/postmortem.md", "nativeTokensPerSec": 57.12, "nativeGeneratedTokens": 7625, "promptTokensPerSec": 2678.82, "usageSource": "engine-log", "durationMs": 181850, "toolCalls": 6, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3618011474609375 }, "peakRssMb": 9980, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-tictactoe", "model": "qwen3.5-4b-q4", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T18:00:38.025Z", "finishedAt": "2026-10-21T18:01:19.916Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear centered board, distinct mark colors and working Restart Game. Alternation, duplicate-click protection, win lock, draw, reset and keyboard input pass without errors or narrow-screen overflow. Readable basic state logic meets the brief; no winning-line emphasis, accessible empty-cell names or live status announcement.", "reviewInputBytes": 4657, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d13578715e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/postmortem.md", "nativeTokensPerSec": 128.13, "nativeGeneratedTokens": 1831, "promptTokensPerSec": 3921.01, "usageSource": "engine-log", "durationMs": 41891, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2659149169921875 }, "peakRssMb": 6534, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-tictactoe", "model": "underdog-saluki-27b-iq2-mix", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T18:02:38.811Z", "finishedAt": "2026-10-21T18:06:09.811Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 5, "functionalCompleteness": 5, "codeQuality": 5, "polish": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear responsive board with colored marks and winning-line feedback. Alternation, repeat-click guard, win lock, draw, reset and keyboard interaction work without errors. Reset removes the highlight. Simple readable logic meets the brief. Body padding adds unnecessary vertical scroll; generic cell labels do not announce changing marks.", "reviewInputBytes": 3970, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ebc2863434.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/postmortem.md", "nativeTokensPerSec": 13.48, "nativeGeneratedTokens": 1895, "promptTokensPerSec": 698.64, "usageSource": "engine-log", "durationMs": 211000, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.286834716796875 }, "peakRssMb": 12232, "peakGpuMemMb": 10709, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-tictactoe", "model": "qwen3.8-27b-q2", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T18:07:27.074Z", "finishedAt": "2026-10-21T18:14:53.982Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "A plain readable board with correct turns, occupied-cell protection, win lock, draw, reset and keyboard operation. No errors or overflow at the inspected widths. State and disabled buttons are synchronized. Monochrome marks, default reset styling, no win highlight and unnamed empty cells limit presentation and accessibility.", "reviewInputBytes": 2741, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1aad1070d3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/postmortem.md", "nativeTokensPerSec": 6, "nativeGeneratedTokens": 2140, "promptTokensPerSec": 489.48, "usageSource": "engine-log", "durationMs": 446908, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2852325439453125 }, "peakRssMb": 14588, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-tictactoe", "model": "gemma4-12b-q4", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T18:16:18.512Z", "finishedAt": "2026-10-21T18:17:19.446Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Centered board, clear status and blue reset control. Alternation, duplicate-click guard, win lock, draw, reset and keyboard activation all work without errors or narrow-screen overflow. Simple code meets the brief. Marks are monochrome, winning cells lack emphasis, empty buttons lack names and status is not announced.", "reviewInputBytes": 4776, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4a7a20716f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/postmortem.md", "nativeTokensPerSec": 140.25, "nativeGeneratedTokens": 2310, "promptTokensPerSec": 1958.16, "usageSource": "engine-log", "durationMs": 60934, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3098297119140625 }, "peakRssMb": 12045, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-tictactoe", "model": "gemma4-e4b-q4", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T18:18:36.732Z", "finishedAt": "2026-10-21T18:20:04.136Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 4, "polish": 3 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The plain board supports turns, repeat-click protection, win lock, draw and keyboard input without errors. Another round requires a page reload; no reset or visible title exists. Game-over logic depends on the word Turn in status text, coupling state to presentation. No winning-line highlight, empty-cell labels or live status.", "reviewInputBytes": 4000, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Targeted interaction checks only. Disabled or pointer-events:none game cells use DOM click() solely to inspect terminal guards. No exhaustive game-tree testing or candidate changes." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/c61eb7689b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/postmortem.md", "nativeTokensPerSec": 191.4, "nativeGeneratedTokens": 6161, "promptTokensPerSec": 4417.97, "usageSource": "engine-log", "durationMs": 87404, "toolCalls": 8, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2877197265625 }, "peakRssMb": 6458, "peakGpuMemMb": 4143, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-petshop", "model": "ornith1.5-9b-q4", "scenario": "petshop", "repetition": 3, "trialId": "petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T18:21:24.433Z", "finishedAt": "2026-10-21T18:29:15.527Z", "success": true, "failureClass": "pass", "reason": "build-me-a-single-page-pet-shop-website/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 1, "functionalCompleteness": 3, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "A local generated logo and service/contact content exist, and the services anchor works. Referenced styles.css is missing, so the1536px logo dominates an unstyled page and forces1544px document width even at390px. The landing page has semantic sections but fails basic visual integration and responsive usability.", "reviewInputBytes": 7298, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "Generated PNG inspected as rendered visual input and separately hashed. Missing stylesheet confirmed; unused logo-generator helper not read/executed. Only local anchor navigation tested." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/postmortem.md", "nativeTokensPerSec": 56.82, "nativeGeneratedTokens": 20890, "promptTokensPerSec": 2598.08, "usageSource": "engine-log", "durationMs": 471094, "toolCalls": 27, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3228759765625 }, "peakRssMb": 13249, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-petshop", "model": "qwen3.5-4b-q4", "scenario": "petshop", "repetition": 3, "trialId": "petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T18:30:42.054Z", "finishedAt": "2026-10-21T18:34:18.653Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on index.html; missing signals stayed unchanged: working-image, image-asset.", "modelEligible": true, "fixedComposite": 3.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 3, "codeQuality": 4, "polish": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable blue service/product layout fits both widths, but the required generated PNG is absent and the header image is broken. An unreferenced SVG does not satisfy the logo contract. The labeled form validates inputs, then falsely claims delivery through a local alert and clears them. Basic styling cannot compensate for missing output and misleading feedback.", "reviewInputBytes": 15886, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity known. Full selected source read and desktop/narrow screenshots inspected; combined20KiB text cap respected.", "Offline installed desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. External HTTP(S) blocked.", "SVG is unreferenced and not substituted for the missing required PNG. Synthetic form submission only exercises the inspected preventDefault/alert/reset handler; no message was sent. Metadata excludes earlier draft content." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/178f97af19.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/postmortem.md", "nativeTokensPerSec": 110.83, "nativeGeneratedTokens": 13202, "promptTokensPerSec": 3672.58, "usageSource": "engine-log", "durationMs": 216599, "toolCalls": 30, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3289031982421875 }, "peakRssMb": 10632, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-petshop", "model": "underdog-saluki-27b-iq2-mix", "scenario": "petshop", "repetition": 3, "trialId": "petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T18:35:19.914Z", "finishedAt": "2026-10-21T19:21:49.475Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"pet-shop-website/workspace/index.html:3:targetnone:frnone:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 7 tool calls (0 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 2.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 1, "functionalCompleteness": 2, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Semantic pet, service and contact sections and working anchors exist, but styles.css, main.js and the required PNG are absent. Both viewports show an unstyled page with a broken logo. Menu does nothing; the footer year is blank. The novalidate form has no handler or delivery path. Readable markup is only an unfinished shell.", "reviewInputBytes": 11723, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded final-source and actual desktop/mobile screenshot review; frozen default-html-v1 integer axes. Final artifact alone receives quality scores, not earlier drafts or tool metadata.", "Offline desktop Chromium at1200x900 and390x844, not mobile-device emulation. Local anchors and Menu button checked; no contact submission, new image generation or candidate edits.", "The form lacks its external script and disables native validation. Its delivery is not implemented in the reviewed files; no network request was sent to test it.", "Compact tool metadata excludes previous source contents and reasoning. Frozen model-stuck classification, deterministic score and runtime measurements remain separate." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/69fc143957.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/postmortem.md", "nativeTokensPerSec": 12.44, "nativeGeneratedTokens": 22823, "promptTokensPerSec": 644.61, "usageSource": "engine-log", "durationMs": 2789561, "toolCalls": 27, "redFlags": [], "autoAnswers": 0, "missingRoles": [ "image-generator" ], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.295501708984375 }, "peakRssMb": 18632, "peakGpuMemMb": 10713, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-petshop", "model": "qwen3.8-27b-q2", "scenario": "petshop", "repetition": 3, "trialId": "petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T19:24:09.760Z", "finishedAt": "2026-10-21T20:48:19.839Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on index.html; missing signals stayed unchanged: working-image, image-asset.", "modelEligible": true, "fixedComposite": 4.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 2, "codeQuality": 3, "polish": 3 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Warm service/contact layout, but the required logo is missing and the advertised product shelf is empty: its script is only a placeholder. The fixed header overflows390px to422px. Internal anchors resolve, yet prominent empty content and a broken brand image leave the page incomplete.", "reviewInputBytes": 7776, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full final-source and rendered-image review using frozen default-html-v1 integer axes and combined20KiB textual cap.", "Offline desktop Chromium1200x900 and390x844, not mobile-device emulation. Targeted interactions only; no candidate edits, new assets or official result changes.", "Earlier drafts and reasoning excluded. Tool metadata preserves routing and repair facts; only the final artifact is scored." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed737b6d12.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/postmortem.md", "nativeTokensPerSec": 4.94, "nativeGeneratedTokens": 23813, "promptTokensPerSec": 473.81, "usageSource": "engine-log", "durationMs": 5050079, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4121551513671875 }, "peakRssMb": 18286, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-petshop", "model": "gemma4-12b-q4", "scenario": "petshop", "repetition": 3, "trialId": "petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T20:50:24.065Z", "finishedAt": "2026-10-21T20:59:17.580Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on index.html; missing signals stayed unchanged: working-image, image-asset.", "modelEligible": true, "fixedComposite": 3.2, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 3, "codeQuality": 4, "polish": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable purple page with responsive cards and labeled booking fields. Required PNG is absent. Form validation/reset works but falsely claims receipt and future contact with no delivery; no appointment time or service is collected. The external hero photo was blocked, so its live appearance is unassessed.", "reviewInputBytes": 18936, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full final-source and rendered-image review using frozen default-html-v1 integer axes and combined20KiB textual cap.", "Offline desktop Chromium1200x900 and390x844, not mobile-device emulation. Targeted interactions only; no candidate edits, new assets or official result changes.", "External hero imagery blocked and unassessed. Source handler inspected before fake local form submission using example.invalid; no real message sent. Earlier drafts/reasoning excluded." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/96caa9b603.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/postmortem.md", "nativeTokensPerSec": 121.3, "nativeGeneratedTokens": 26065, "promptTokensPerSec": 1840.42, "usageSource": "engine-log", "durationMs": 533515, "toolCalls": 51, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4221343994140625 }, "peakRssMb": 18529, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-petshop", "model": "gemma4-e4b-q4", "scenario": "petshop", "repetition": 3, "trialId": "petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T21:00:41.036Z", "finishedAt": "2026-10-21T21:03:02.236Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website-build/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 4, "codeQuality": 4, "polish": 3 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Generated PNG and four product cards load, and cart alerts clearly disclose simulation. About and Contact navigation target sections that do not exist. Large product-image placeholders, garbled logo lettering and sparse content limit finish. The narrow layout has a minor1px overflow; it remains readable.", "reviewInputBytes": 4570, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full final-source and rendered-image review using frozen default-html-v1 integer axes and combined20KiB textual cap.", "Offline desktop Chromium1200x900 and390x844, not mobile-device emulation. Targeted interactions only; no candidate edits, new assets or official result changes.", "Generated logo evaluated as rendered visual input; placeholder product boxes are part of the delivered page. Local cart alert checked; no purchase or external communication." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/824365ad63.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/postmortem.md", "nativeTokensPerSec": 173.45, "nativeGeneratedTokens": 8724, "promptTokensPerSec": 4514.74, "usageSource": "engine-log", "durationMs": 141200, "toolCalls": 17, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3303375244140625 }, "peakRssMb": 7227, "peakGpuMemMb": 6715, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-petshop", "model": "qwen3.5-9b-q4", "scenario": "petshop", "repetition": 3, "trialId": "petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 2, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T21:04:46.650Z", "finishedAt": "2026-10-21T21:09:23.092Z", "success": true, "failureClass": "pass", "reason": "pet-shop-website-with-logo/workspace/index.html passed sniff (signals: pet-vocab, store-vocab, structured-page, working-image, image-asset) + runtime (1 assertion(s))", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 4, "codeQuality": 4, "polish": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Generated logo, service sections and working anchors fit both widths. Form labels and native validation work, but the local-only alert promises a reply without sending anything. About/contact headings sit flush against the edge with uneven spacing. Clear basic page, with unfinished presentation and misleading submission feedback.", "reviewInputBytes": 5322, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full final-source and rendered-image review using frozen default-html-v1 integer axes and combined20KiB textual cap.", "Offline desktop Chromium1200x900 and390x844, not mobile-device emulation. Targeted interactions only; no candidate edits, new assets or official result changes.", "Source handler inspected before fake local form submission using example.invalid; no real message sent. Generated logo evaluated as rendered visual input." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/77ace886ab.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/postmortem.md", "nativeTokensPerSec": 85.42, "nativeGeneratedTokens": 9462, "promptTokensPerSec": 2478.01, "usageSource": "engine-log", "durationMs": 276442, "toolCalls": 29, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.364654541015625 }, "peakRssMb": 13267, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-tankcombat", "model": "qwen3.5-4b-q4", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T21:11:09.304Z", "finishedAt": "2026-10-21T21:22:17.599Z", "success": true, "failureClass": "pass", "reason": "tank-combat-arcade-4/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))", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 1, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Movement and shooting respond, but enemies remain above the canvas: update increments dy without changing y. Combat never develops in replay. Destroyed enemies also are not removed in source. Bare grid, no initial instructions and a clipped fixed canvas undermine usability. Passing input checks do not establish a playable game.", "reviewInputBytes": 8017, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full final-source and rendered-image review using frozen default-html-v1 integer axes and combined20KiB textual cap.", "Offline desktop Chromium1200x900 and390x844, not mobile-device emulation. Targeted interactions only; no candidate edits, new assets or official result changes.", "Arrow/D/Space inputs and6.5s natural gameplay inspected with read-only state probes. No injected state or forced hits; enemy-position defect also confirmed in source. No scored kill/game-over observed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1da550bc0b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/postmortem.md", "nativeTokensPerSec": 126.16, "nativeGeneratedTokens": 57401, "promptTokensPerSec": 3751.37, "usageSource": "engine-log", "durationMs": 668295, "toolCalls": 23, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.30291748046875 }, "peakRssMb": 13679, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-tankcombat", "model": "underdog-saluki-27b-iq2-mix", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T21:24:38.139Z", "finishedAt": "2026-10-21T21:32:47.000Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 2, "codeQuality": 2, "polish": 2 }, "advisoryMean": 2.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear scaled arena, pursuing enemies, Arrow movement and R reset work. WASD and Space fail because code-key names are indexed with event.key. Millisecond cooldown/respawn constants are decremented in seconds: a hit leaves roughly900s of respawn delay. Lives drop to2 while HUD stays3. Combat is incomplete despite its passing gate.", "reviewInputBytes": 8583, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full final-source and rendered-image review using frozen default-html-v1 integer axes and combined20KiB textual cap.", "Offline desktop Chromium1200x900 and390x844, not mobile-device emulation. Targeted interactions only; no candidate edits, new assets or official result changes.", "Ordinary Arrow/D/Space/R inputs and natural enemy motion inspected. Keyboard steering caused a real collision without state injection; read-only probes confirmed staleHUD and900→~899s respawn timer. No scored kill or full game-over observed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4f43c38a23.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/postmortem.md", "nativeTokensPerSec": 12.58, "nativeGeneratedTokens": 5033, "promptTokensPerSec": 649.48, "usageSource": "engine-log", "durationMs": 488861, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.294525146484375 }, "peakRssMb": 13588, "peakGpuMemMb": 10709, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-tankcombat", "model": "qwen3.8-27b-q2", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T21:34:19.385Z", "finishedAt": "2026-10-21T21:54:29.274Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 4, "polish": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable grid, HUD and instructions. Arrow/WASD movement, visible Space shots, enemy fire, natural game-over and R restart work. Source implements scored kills and new waves, but neither was observed in short play. Fixed canvas clips at narrow widths; frame-based motion and sparse hit feedback limit robustness and finish.", "reviewInputBytes": 7235, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full-source review with frozen integer axes and 20 KiB combined text cap.", "Offline desktop Chromium at 1200x900 and 390x844; ordinary keyboard inputs, read-only state inspection and rendered images. No injected game state or candidate changes.", "Natural death and R restart observed; no scored kill or wave transition observed. Short random gameplay is not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/a2ee0048c9.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/postmortem.md", "nativeTokensPerSec": 5.94, "nativeGeneratedTokens": 6621, "promptTokensPerSec": 493.67, "usageSource": "engine-log", "durationMs": 1209889, "toolCalls": 2, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2916107177734375 }, "peakRssMb": 15577, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-tankcombat", "model": "gemma4-12b-q4", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T21:56:14.917Z", "finishedAt": "2026-10-21T21:58:18.288Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 2, "codeQuality": 2, "polish": 2 }, "advisoryMean": 2.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive arena and visible controls, but W never moves the tank: acceleration times friction is below the cutoff and speed resets to zero each frame. D rotates and Space fires, though shots travel 90 degrees from the drawn barrel. Enemies pursue; collision triggers a game-over alert and reload. The player is stranded despite the passing gate.", "reviewInputBytes": 7616, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full-source review with frozen integer axes and 20 KiB combined text cap.", "Offline desktop Chromium at 1200x900 and 390x844; ordinary keyboard inputs, read-only lexical state inspection and rendered images. No injected game state or candidate changes.", "W held 500 ms leaves position and speed unchanged. Natural collision, alert and reload observed; no scored kill observed. Reverse movement has the same cutoff defect in source but was not separately replayed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2f94943bbb.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/postmortem.md", "nativeTokensPerSec": 143.49, "nativeGeneratedTokens": 6444, "promptTokensPerSec": 1875.99, "usageSource": "engine-log", "durationMs": 123371, "toolCalls": 7, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.31011962890625 }, "peakRssMb": 17517, "peakGpuMemMb": 8705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-tankcombat", "model": "gemma4-e4b-q4", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T21:59:53.812Z", "finishedAt": "2026-10-21T22:00:58.873Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 1, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Movement, shots and enemy pursuit work, but collision scoring never runs: updateProjectiles receives two arguments while target is the third. Shots through an overlapping enemy leave score zero. Enemy damage/fire and game-over are unfinished; enemy rotation is undefined. Bare boxes, no instructions and a clipped fixed canvas limit usability.", "reviewInputBytes": 8567, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full-source review with frozen integer axes and 20 KiB combined text cap.", "Offline desktop Chromium at 1200x900 and 390x844; ordinary keyboard inputs, read-only lexical state inspection and rendered images. No injected game state or candidate changes.", "Eight real Space presses while enemy overlaps player produce projectiles without damage or score. No win/loss/restart implementation. The idle input counter advances without input, so its passing signature does not certify gameplay." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2bfa12878f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/postmortem.md", "nativeTokensPerSec": 209.15, "nativeGeneratedTokens": 5192, "promptTokensPerSec": 4480.32, "usageSource": "engine-log", "durationMs": 65061, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.401763916015625 }, "peakRssMb": 6218, "peakGpuMemMb": 4143, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-tankcombat", "model": "qwen3.5-9b-q4", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T22:02:21.879Z", "finishedAt": "2026-10-21T22:17:50.966Z", "success": false, "failureClass": "model", "reason": "retry loop (fast-path): sniff \"tank-combat-arcade/workspace/index.html:9:targetnone:frnone:rp2:rf1\" stuck for 8m — 4 re-writes (8 tool calls) without sniff movement. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 2, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive but bare arena. Movement is only 3 px/s and projectiles 6 px/s; shots always travel right because player angle never changes. Arrow/WASD and Space do respond, but the requested window/dataset input hook is absent. No instructions or heading cue. Combat is impractically slow; no scored kill or game-over was observed.", "reviewInputBytes": 8404, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded full-source review with frozen integer axes and 20 KiB combined text cap including diagnostic metadata.", "Offline desktop Chromium at 1200x900 and 390x844; ordinary keyboard inputs, read-only lexical state inspection and rendered images. No injected game state or candidate changes.", "No page errors in this final artifact replay. Earlier keys-is-not-defined failures are log history, not attributed to the final source. The explicit observability contract is unmet despite actual input response; official failure retained. Short gameplay is not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/postmortem.md", "nativeTokensPerSec": 78.97, "nativeGeneratedTokens": 9536, "promptTokensPerSec": 2487.29, "usageSource": "engine-log", "durationMs": 929087, "toolCalls": 28, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4049224853515625 }, "peakRssMb": 15425, "peakGpuMemMb": 8185, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-tankcombat", "model": "ornith1.5-9b-q4", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T22:19:47.432Z", "finishedAt": "2026-10-21T22:41:10.589Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 7.5, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 2, "codeQuality": 2, "polish": 3 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Attractive start screen, HUD and pause/resume. Vertical movement and Space shots work, but no enemies ever spawn: start clears the array and the only spawn call depends on defeating an enemy. The barrel stays right-facing while shots rotate; horizontal translation is absent. Fixed canvas clips on narrow screens. Combat never begins.", "reviewInputBytes": 15943, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen HTML rubric, full final source below 20 KiB. Desktop, after-input and narrow rendered images viewed.", "Offline Chromium at 1200x900 and 390x844; ordinary Start/Arrow/Space/P/Resume controls. No state injection. IIFE state is private; source and visible empty playfield establish the missing spawn path.", "No scored kill, damage or game-over observed; these paths cannot begin normally without an enemy spawn. No candidate change or official result reclassification." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/874af04527.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/postmortem.md", "nativeTokensPerSec": 55.8, "nativeGeneratedTokens": 51699, "promptTokensPerSec": 2386.53, "usageSource": "engine-log", "durationMs": 1283157, "toolCalls": 25, "redFlags": [ { "gezel": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/sessions/wren", "tool": "(no write_file this turn)", "argsSummary": "content_chars=2411 h2_count=3 preview=\"I've hit a blocker and want to stop retrying it rather than loop. ## Status **Project created:** `Tank Combat Arcade` …\"", "atTurn": 2, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." } ], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4667205810546875 }, "peakRssMb": 11702, "peakGpuMemMb": 7339, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-schema-migration", "model": "underdog-saluki-27b-iq2-mix", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T22:44:17.054Z", "finishedAt": "2026-10-21T22:50:32.350Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Types, store and handlers are coherently migrated; required tests and a clear guide are present. Six name-split probes pass without mutation. However, the function drops extra non-name fields despite the preservation requirement and the guide explicitly promising to keep them. Configurations remain unchanged.", "reviewInputBytes": 10140, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/1174d9dcd5.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/postmortem.md", "nativeTokensPerSec": 13.02, "nativeGeneratedTokens": 2950, "promptTokensPerSec": 662.69, "usageSource": "engine-log", "durationMs": 375296, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.331085205078125 }, "peakRssMb": 17111, "peakGpuMemMb": 10723, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-schema-migration", "model": "qwen3.8-27b-q2", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T22:51:36.579Z", "finishedAt": "2026-10-21T23:02:13.494Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Complete focused migration with matching tests and rollout guidance. Six split cases preserve spacing, id/email and extra fields without mutating input; the old name is removed. Store and handlers work, strict source typing is clean, and the virtual consumer correctly rejects user.name. Configuration is unchanged.", "reviewInputBytes": 10030, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/e232744039.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/postmortem.md", "nativeTokensPerSec": 6.87, "nativeGeneratedTokens": 3018, "promptTokensPerSec": 472.6, "usageSource": "engine-log", "durationMs": 636915, "toolCalls": 20, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.339019775390625 }, "peakRssMb": 17988, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-schema-migration", "model": "gemma4-12b-q4", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T23:03:19.914Z", "finishedAt": "2026-10-21T23:05:39.353Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Concise working migration, updated store/handlers and three meaningful full-record tests. All six split cases pass, but only id/email are carried through: extra fields are dropped despite the guide promising preservation. Comments still describe the old shape. Package and strict configuration are unchanged.", "reviewInputBytes": 9144, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/0732c7ad7e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/postmortem.md", "nativeTokensPerSec": 120.38, "nativeGeneratedTokens": 3850, "promptTokensPerSec": 1886.48, "usageSource": "engine-log", "durationMs": 139439, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4035491943359375 }, "peakRssMb": 16939, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-schema-migration", "model": "gemma4-e4b-q4", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T23:06:42.983Z", "finishedAt": "2026-10-21T23:09:08.822Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 3, "scopeDiscipline": 4 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core types, store, handlers and five tests are present. Migration trims and normalizes whitespace instead of preserving the suffix after the first space; two edge probes fail. It also drops extra fields and makes legacy email optional, contradicting preservation claims. The guide is verbose and overstates fidelity.", "reviewInputBytes": 12635, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9773a858d9.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/postmortem.md", "nativeTokensPerSec": 191.38, "nativeGeneratedTokens": 9689, "promptTokensPerSec": 4486.14, "usageSource": "engine-log", "durationMs": 145839, "toolCalls": 23, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.43182373046875 }, "peakRssMb": 6455, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-schema-migration", "model": "qwen3.5-9b-q4", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T23:10:03.133Z", "finishedAt": "2026-10-21T23:12:19.934Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Actual migration passes six split probes and preserves id/email, with matching tests and updated consumers. Extra fields are dropped. The guide embeds a different split(name,2) implementation that would truncate multi-word surnames, and suggests nonempty last names despite the one-word case. Production code is stronger than its guidance.", "reviewInputBytes": 10272, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/128ff317ff.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/postmortem.md", "nativeTokensPerSec": 85.47, "nativeGeneratedTokens": 3810, "promptTokensPerSec": 2539.75, "usageSource": "engine-log", "durationMs": 136801, "toolCalls": 21, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4485015869140625 }, "peakRssMb": 13006, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-schema-migration", "model": "ornith1.5-9b-q4", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T23:13:14.191Z", "finishedAt": "2026-10-21T23:15:57.880Z", "success": true, "failureClass": "pass", "reason": "all 7 signals firing including tsc-clean (signals: types-updated, store-updated, handlers-updated, migrate-function, tests-present, migration-doc, tsc-clean)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 4, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Six split cases and the migrated consumers work; tests and guide cover the required shapes. Extra fields are discarded despite the broad preservation promise. A separate MigratedUser duplicates the canonical type, and old handler comments remain. Strict source typing passes; configuration and fixture files are preserved.", "reviewInputBytes": 10119, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/dc4a2129a8.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/postmortem.md", "nativeTokensPerSec": 61.62, "nativeGeneratedTokens": 4704, "promptTokensPerSec": 2747.62, "usageSource": "engine-log", "durationMs": 163689, "toolCalls": 26, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.390869140625 }, "peakRssMb": 11046, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-schema-migration", "model": "qwen3.5-4b-q4", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T23:16:51.852Z", "finishedAt": "2026-10-21T23:18:04.077Z", "success": false, "failureClass": "model", "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 runtime: multi-word: wrong name split (expected {\"firstName\":\"Mary\",\"lastName\":\"Anne Smith\"}, got {\"firstName\":\"Mary\",\"lastName\":\"Anne\"})", "modelEligible": true, "fixedComposite": 4.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 5 }, "advisoryMean": 3.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Types, store and handlers are migrated, but split(name,2) discards words after the second token and loses repeated-space suffixes. Tests and MIGRATION.md are absent. Source typing passes, yet two of six preservation probes fail and extra fields are dropped. The package and strict configuration remain untouched.", "reviewInputBytes": 5259, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production migration/store/handler exports were probed with ordinary inputs; source and a virtual consumer compiled without modifying the workspace. Authored tests were read, not executed. Extra fields extend the frozen id/email gate and test the broader preservation wording and documentation claims.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/be5486180b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/postmortem.md", "nativeTokensPerSec": 145.17, "nativeGeneratedTokens": 1161, "promptTokensPerSec": 4056.35, "usageSource": "engine-log", "durationMs": 72225, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.314300537109375 }, "peakRssMb": 8210, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-failing-tests-spec", "model": "qwen3.8-27b-q2", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-21T23:18:58.378Z", "finishedAt": "2026-10-21T23:23:42.441Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear state table, guarded expedite and correct history; all 84 ordinary state/event combinations pass. Inherited property names are accepted as events: constructor/toString turn state into a function and __proto__ into an object. All 36 such probes fail. Tests and configuration remain unchanged.", "reviewInputBytes": 7054, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production API only: 120 state/event/guard cases, legal setup transitions and separate default construction. Setup failures are retained, not bypassed with injected state. Extra unknown-string events extend the frozen tests; no official outcome is replaced.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/e8824696b4.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/postmortem.md", "nativeTokensPerSec": 7.94, "nativeGeneratedTokens": 860, "promptTokensPerSec": 503.94, "usageSource": "engine-log", "durationMs": 284063, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.31317138671875 }, "peakRssMb": 18476, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-failing-tests-spec", "model": "gemma4-12b-q4", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T23:24:43.985Z", "finishedAt": "2026-10-21T23:25:18.950Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 84 ordinary transition, guard and history checks pass, with untouched tests/configuration. Duplicated can/send checks are readable but accept inherited object-property events; constructor, toString and __proto__ set state to undefined. All 36 special-name cases fail, exposing a public string-input robustness gap.", "reviewInputBytes": 7611, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production API only: 120 state/event/guard cases, legal setup transitions and separate default construction. Setup failures are retained, not bypassed with injected state. Extra unknown-string events extend the frozen tests; no official outcome is replaced.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/e95d204c40.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/postmortem.md", "nativeTokensPerSec": 138.36, "nativeGeneratedTokens": 681, "promptTokensPerSec": 2024, "usageSource": "engine-log", "durationMs": 34965, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.256134033203125 }, "peakRssMb": 11498, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-failing-tests-spec", "model": "gemma4-e4b-q4", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T23:26:12.650Z", "finishedAt": "2026-10-21T23:35:06.906Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Default construction, all 84 ordinary cases and all 36 unknown-property-name cases pass. Explicit membership checks avoid inherited-event corruption and the history getter copies its array. Two transition structures and extensive comments add maintenance overhead; exported readonly qualifiers were removed. Tests and configuration remain unchanged.", "reviewInputBytes": 8624, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production API only: 120 state/event/guard cases, legal setup transitions and separate default construction. Setup failures are retained, not bypassed with injected state. Extra unknown-string events extend the frozen tests; no official outcome is replaced.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/adc1f30dc0.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/postmortem.md", "nativeTokensPerSec": 221.68, "nativeGeneratedTokens": 26409, "promptTokensPerSec": 4314.45, "usageSource": "engine-log", "durationMs": 534256, "toolCalls": 39, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 2, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3137054443359375 }, "peakRssMb": 12852, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-failing-tests-spec", "model": "qwen3.5-9b-q4", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T23:36:01.249Z", "finishedAt": "2026-10-21T23:37:59.494Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Focused implementation passes all 84 ordinary transition/guard/history cases. Public string events still reach inherited table properties: constructor/toString change state to a function and __proto__ to an object; all 36 special-name cases fail. Source types are clean and tests/configuration are unchanged.", "reviewInputBytes": 7287, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production API only: 120 state/event/guard cases, legal setup transitions and separate default construction. Setup failures are retained, not bypassed with injected state. Extra unknown-string events extend the frozen tests; no official outcome is replaced.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/3e41108a46.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/postmortem.md", "nativeTokensPerSec": 95.94, "nativeGeneratedTokens": 2330, "promptTokensPerSec": 2564.23, "usageSource": "engine-log", "durationMs": 118245, "toolCalls": 16, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3043670654296875 }, "peakRssMb": 12485, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-failing-tests-spec", "model": "ornith1.5-9b-q4", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T23:38:53.724Z", "finishedAt": "2026-10-21T23:40:50.796Z", "success": false, "failureClass": "model", "reason": "repair-aborted: priya/13468341 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: tests/machine.test.ts", "modelEligible": true, "fixedComposite": 5.6, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 1, "codeQuality": 3, "robustness": 1, "scopeDiscipline": 5 }, "advisoryMean": 2.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Starts in paid instead of draft, returns a stale state snapshot and ignores the expedite option. Legal setup paths cannot be reached, so none of the 120 probes pass; those counts share setup failures rather than 120 independent bugs. Types compile and tests/configuration are untouched, but the basic API contract is broken.", "reviewInputBytes": 12662, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production API only: 120 state/event/guard cases, legal setup transitions and separate default construction. Setup failures are retained, not bypassed with injected state. Extra unknown-string events extend the frozen tests; no official outcome is replaced.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive.", "Recorded tools show prior successful spec reads, then repeated machine.ts reads during the final request for tests/machine.test.ts. Preserve repair-aborted/model-default; do not claim the spec was never read or call this GPU failure." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/d82ea8785b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/postmortem.md", "nativeTokensPerSec": 64.59, "nativeGeneratedTokens": 1742, "promptTokensPerSec": 2767.22, "usageSource": "engine-log", "durationMs": 117072, "toolCalls": 25, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4254913330078125 }, "peakRssMb": 11932, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-failing-tests-spec", "model": "qwen3.5-4b-q4", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T23:42:35.117Z", "finishedAt": "2026-10-21T23:43:15.203Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 4, "codeQuality": 5, "robustness": 4, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Readable switch-based implementation rejects unknown events and preserves history. However, expedited machines reject normal ship from paid, though expedite is an additional option. This also prevents several probe setup paths; it is one root defect, not many independent failures. Tests/configuration remain unchanged and source typing is clean.", "reviewInputBytes": 7378, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded review using frozen campaign-code-v1 integer axes; full selected text and diagnostic grounding are under the combined 20 KiB cap. Only produced/changed artifacts are scored.", "Production API only: 120 state/event/guard cases, legal setup transitions and separate default construction. Setup failures are retained, not bypassed with injected state. Extra unknown-string events extend the frozen tests; no official outcome is replaced.", "Selected files hashed. No candidate edits, dependency installation, model rerun, fixture weakening or gate changes. Tests are targeted and not exhaustive." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/babbbf2f7c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/postmortem.md", "nativeTokensPerSec": 130.5, "nativeGeneratedTokens": 648, "promptTokensPerSec": 3803.78, "usageSource": "engine-log", "durationMs": 40086, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.244140625 }, "peakRssMb": 5900, "peakGpuMemMb": 5801, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec", "model": "underdog-saluki-27b-iq2-mix", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T23:44:14.997Z", "finishedAt": "2026-10-21T23:45:55.002Z", "success": true, "failureClass": "pass", "reason": "vitest and strict tsc gates green with the seeded spec untouched (signals: tests-untouched, machine-implemented, vitest-green, tsc-clean)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 5, "robustness": 3, "scopeDiscipline": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear table-based implementation with default draft state, correct guards and history; all 84 ordinary combinations pass. However, inherited object names are accepted by in/direct lookup: constructor and toString turn state into a function, while __proto__ turns it into an object. All 36 special-name probes fail. Tests/configuration are unchanged.", "reviewInputBytes": 6995, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 integer axes; full source and grounding under the combined 20 KiB cap.", "Only inspected production exports executed through ordinary API calls; 120 state/event/guard cases and default construction checked. No state injection or candidate test execution.", "Additional unknown-string events are advisory and do not replace the frozen gate. No candidate edits, installation, inference rerun or official score change." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/7884b5e425.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/postmortem.md", "nativeTokensPerSec": 12.97, "nativeGeneratedTokens": 551, "promptTokensPerSec": 690.39, "usageSource": "engine-log", "durationMs": 100005, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.228515625 }, "peakRssMb": 11273, "peakGpuMemMb": 10705, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-symptom-debug", "model": "gemma4-12b-q4", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-21T23:46:53.262Z", "finishedAt": "2026-10-21T23:47:46.414Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Focused correction removes the extra cursor increment and updates the comment to describe resuming at the cursor. All 62 targeted pagination, invalid-input and direct-index checks pass without input mutation. The acceptance script and package remain unchanged; unrelated behavior is preserved.", "reviewInputBytes": 4147, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; complete selected text below the combined 20 KiB cap.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. These checks are not exhaustive.", "Acceptance script and package verified unchanged. No official gate rerun, candidate edits, installation or runtime changes. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/2d56e46bf3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/postmortem.md", "nativeTokensPerSec": 129.3, "nativeGeneratedTokens": 2408, "promptTokensPerSec": 1693.15, "usageSource": "engine-log", "durationMs": 53152, "toolCalls": 7, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2603302001953125 }, "peakRssMb": 13269, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-symptom-debug", "model": "gemma4-e4b-q4", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-21T23:48:57.032Z", "finishedAt": "2026-10-21T23:49:30.403Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The focused one-line correction removes the cursor skip and passes all 62 targeted checks without input mutation. However, the adjacent comment still incorrectly says one past the cursor, and the edited line loses indentation. Acceptance script and package remain unchanged.", "reviewInputBytes": 4149, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; complete selected text below the combined 20 KiB cap.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. These checks are not exhaustive.", "Acceptance script and package verified unchanged. No official gate rerun, candidate edits, installation or runtime changes. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/28991eb661.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/postmortem.md", "nativeTokensPerSec": 188.34, "nativeGeneratedTokens": 2130, "promptTokensPerSec": 4594.26, "usageSource": "engine-log", "durationMs": 33371, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.247833251953125 }, "peakRssMb": 4486, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-symptom-debug", "model": "qwen3.5-9b-q4", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-21T23:50:28.315Z", "finishedAt": "2026-10-21T23:51:14.222Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line cursor correction passes all 62 targeted pagination, invalid-input and direct-index checks without mutating input. The adjacent comment still incorrectly describes resuming one past the cursor. The acceptance script and package are unchanged, and the change stays within scope.", "reviewInputBytes": 4151, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; complete selected text below the combined 20 KiB cap.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. These checks are not exhaustive.", "Acceptance script and package verified unchanged. No official gate rerun, candidate edits, installation or runtime changes. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/35723d5cb4.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/postmortem.md", "nativeTokensPerSec": 64.33, "nativeGeneratedTokens": 1234, "promptTokensPerSec": 2530.73, "usageSource": "engine-log", "durationMs": 45907, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2476654052734375 }, "peakRssMb": 7355, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-symptom-debug", "model": "ornith1.5-9b-q4", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-21T23:52:12.934Z", "finishedAt": "2026-10-21T23:53:21.615Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line correction removes the cursor skip and passes all 62 targeted checks without input mutation. The adjacent comment still incorrectly describes resuming one past the cursor. Acceptance script and package are unchanged; the fix stays within scope.", "reviewInputBytes": 4151, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; full final module, unchanged acceptance script and original module grounding were read within the combined 20 KiB cap. Only the correction is scored.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. Checks include input nonmutation and are not exhaustive.", "Acceptance script/package seed equality verified; no official gate rerun, candidate edits or installation. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9843bc717e.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/postmortem.md", "nativeTokensPerSec": 56.61, "nativeGeneratedTokens": 2345, "promptTokensPerSec": 2642.77, "usageSource": "engine-log", "durationMs": 68681, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2430877685546875 }, "peakRssMb": 6724, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-symptom-debug", "model": "qwen3.5-4b-q4", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-21T23:54:20.832Z", "finishedAt": "2026-10-21T23:55:39.222Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line correction removes the cursor skip and passes all 62 targeted checks without input mutation. The adjacent comment still incorrectly describes resuming one past the cursor. Acceptance script and package are unchanged; the fix stays within scope.", "reviewInputBytes": 4151, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; full final module, unchanged acceptance script and original module grounding were read within the combined 20 KiB cap. Only the correction is scored.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. Checks include input nonmutation and are not exhaustive.", "Acceptance script/package seed equality verified; no official gate rerun, candidate edits or installation. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/beeea1ff13.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/postmortem.md", "nativeTokensPerSec": 102.39, "nativeGeneratedTokens": 4055, "promptTokensPerSec": 3828.53, "usageSource": "engine-log", "durationMs": 78390, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2939605712890625 }, "peakRssMb": 7795, "peakGpuMemMb": 5801, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-symptom-debug", "model": "underdog-saluki-27b-iq2-mix", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-21T23:56:33.612Z", "finishedAt": "2026-10-22T00:00:57.134Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Focused correction removes the extra cursor increment and updates the comment to explain the next item boundary. All 62 targeted pagination, invalid-input and direct-index checks pass without input mutation. Acceptance script and package remain unchanged; unrelated behavior is preserved.", "reviewInputBytes": 4169, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; full final module, unchanged acceptance script and original module grounding were read within the combined 20 KiB cap. Only the correction is scored.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. Checks include input nonmutation and are not exhaustive.", "Acceptance script/package seed equality verified; no official gate rerun, candidate edits or installation. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/5e1899dfc5.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/postmortem.md", "nativeTokensPerSec": 12.59, "nativeGeneratedTokens": 2673, "promptTokensPerSec": 692.9, "usageSource": "engine-log", "durationMs": 263522, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.23272705078125 }, "peakRssMb": 9860, "peakGpuMemMb": 10707, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-symptom-debug", "model": "qwen3.8-27b-q2", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-22T00:01:51.458Z", "finishedAt": "2026-10-22T00:05:28.419Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 6, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 5.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Focused correction removes the extra cursor increment and updates the comment to explain the next page boundary. All 62 targeted pagination, invalid-input and direct-index checks pass without input mutation. Acceptance script and package remain unchanged; unrelated behavior is preserved.", "reviewInputBytes": 4173, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-code-v1 review; full final module, unchanged acceptance script and original module grounding were read within the combined 20 KiB cap. Only the correction is scored.", "Only inspected production exports executed: 48 varied-size walks, six invalid limits, three malformed cursors and five direct-index cases. Checks include input nonmutation and are not exhaustive.", "Acceptance script/package seed equality verified; no official gate rerun, candidate edits or installation. Inherited cursor-parser permissiveness was not broadly audited." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/72de12e1d3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/postmortem.md", "nativeTokensPerSec": 4.95, "nativeGeneratedTokens": 752, "promptTokensPerSec": 510.99, "usageSource": "engine-log", "durationMs": 216961, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2337646484375 }, "peakRssMb": 11768, "peakGpuMemMb": 11681, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-data-wrangle", "model": "gemma4-e4b-q4", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-22T00:06:24.260Z", "finishedAt": "2026-10-22T00:09:16.641Z", "success": false, "failureClass": "model", "reason": "repair-exhausted (score plateau): 6 completed repairs with the scenario score frozen at 4 while the failing detail kept changing; latest checked file out/customers.json. Last failure: email \"nils.okafor@example.com\" appears more than once — dedup is missing", "modelEligible": true, "fixedComposite": 4.4, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 2, "completeness": 3, "traceability": 2, "outputUsability": 2 }, "advisoryMean": 2.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Valid JSON has 52 rows for only 40 of 56 customers, with 12 duplicate occurrences, seven field mismatches and 26 sorting inversions. Some ids are assigned to the wrong customers and Zoe's name becomes an email. The selected .mjs helper uses require and a broken header split. Raw inputs/checker are intact; the supplied helper cannot reproduce a valid transformation.", "reviewInputBytes": 17789, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 review; complete final JSON, explicitly selected latest authored helper, unchanged checker and three raw CSVs read. Compact actual tool metadata and selected terminal log also count toward the combined 20 KiB cap.", "Independent reference normalization derives from unchanged frozen CSV seeds. Candidate helpers were read but not executed or repaired; source-level defects are distinguished from actual delivered-data mismatches.", "Earlier helper bodies/rejected writes/reasoning were excluded before reading. Tool error excerpts are bounded; they do not establish which other tools were available on every turn. Official classifications, scores and runtime settings remain unchanged." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8de5cf5028.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/postmortem.md", "nativeTokensPerSec": 213.75, "nativeGeneratedTokens": 24413, "promptTokensPerSec": 4537.43, "usageSource": "engine-log", "durationMs": 172381, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3387298583984375 }, "peakRssMb": 6106, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-data-wrangle", "model": "qwen3.5-9b-q4", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-22T00:10:11.193Z", "finishedAt": "2026-10-22T00:25:36.527Z", "success": false, "failureClass": "model", "reason": "retry loop (fast-path): sniff \"data-wrangle:6:targetnone:fr1hgloej:rp0:rf0\" stuck for 10m — 3 re-writes (11 tool calls) without sniff movement. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 5.1, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 4, "completeness": 4, "traceability": 3, "outputUsability": 4 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The JSON has 54 unique customers, missing Leo and Wim; Omar's date is wrong and eight sorting inversions remain. Names/source ids otherwise match the reference. Its helper mishandles quoted CSV fields and emits original date strings rather than normalized dates, so it does not reproduce the supplied JSON. Raw inputs/checker are unchanged.", "reviewInputBytes": 19434, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 review; complete final JSON, explicitly selected latest authored helper, unchanged checker and three raw CSVs read. Compact actual tool metadata and selected terminal log also count toward the combined 20 KiB cap.", "Independent reference normalization derives from unchanged frozen CSV seeds. Candidate helpers were read but not executed or repaired; source-level defects are distinguished from actual delivered-data mismatches.", "Earlier helper bodies/rejected writes/reasoning were excluded before reading. Tool error excerpts are bounded; they do not establish which other tools were available on every turn. Official classifications, scores and runtime settings remain unchanged." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8297171e93.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/postmortem.md", "nativeTokensPerSec": 83.81, "nativeGeneratedTokens": 29397, "promptTokensPerSec": 2220.74, "usageSource": "engine-log", "durationMs": 925334, "toolCalls": 26, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3268890380859375 }, "peakRssMb": 12384, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-data-wrangle", "model": "ornith1.5-9b-q4", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-22T00:28:04.120Z", "finishedAt": "2026-10-22T00:56:10.362Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"data-wrangle:7:targetnone:frexk4xx:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 19 tool calls (1 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 4.8, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 5, "completeness": 5, "traceability": 5, "outputUsability": 4 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 unique customers and every field match the normalized source, but seven adjacent email pairs are unsorted. The readable cleaner produces the exact sorted reference in an isolated unchanged-code replay, showing a delivery/execution gap. Raw inputs/checker are unchanged. Replay success does not replace the final unsorted artifact.", "reviewInputBytes": 20366, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 review with the combined 20 KiB selected-text cap, including unchanged grounding and selected diagnostics. Only model-produced output/helpers are scored.", "Independent reference normalization uses unchanged frozen CSVs. The already-inspected main helper was replayed without modification in a fresh copy containing the same CSVs and an existing empty out directory. Replay results derive solely from selected source and are not substitute scored artifacts.", "No original candidate changes, dependency installation, model rerun or official gate rerun. Earlier drafts/reasoning excluded. Runtime tool metadata does not prove which alternative tools were available on each turn.", "The replay checks supplied CSVs only, not arbitrary malformed data or a broad CSV conformance suite. Ornith tmp/run.js was inspected but not separately executed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/62fd0bcf50.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/postmortem.md", "nativeTokensPerSec": 55.39, "nativeGeneratedTokens": 78876, "promptTokensPerSec": 2060.35, "usageSource": "engine-log", "durationMs": 1686242, "toolCalls": 31, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2440032958984375 }, "peakRssMb": 8076, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-data-wrangle", "model": "qwen3.5-4b-q4", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-22T00:57:08.899Z", "finishedAt": "2026-10-22T01:01:56.093Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on out/customers.json; missing signals stayed unchanged: row-count. Last failure: expected exactly 56 records after dedup, got 43", "modelEligible": true, "fixedComposite": 5.1, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 3, "traceability": 3, "outputUsability": 3 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Valid JSON contains 43 of 56 customers, retaining stale Alice, Zoe and Robert records with seven field mismatches and 16 sorting inversions. The latest cleaner flattens CSV fields but expects complete rows; unchanged-code replay fails with Missing required column in header and creates no output. Raw inputs/checker are unchanged.", "reviewInputBytes": 16753, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 review with the combined 20 KiB selected-text cap, including unchanged grounding and selected diagnostics. Only model-produced output/helpers are scored.", "Independent reference normalization uses unchanged frozen CSVs. The already-inspected main helper was replayed without modification in a fresh copy containing the same CSVs and an existing empty out directory. Replay results derive solely from selected source and are not substitute scored artifacts.", "No original candidate changes, dependency installation, model rerun or official gate rerun. Earlier drafts/reasoning excluded. Runtime tool metadata does not prove which alternative tools were available on each turn.", "The replay checks supplied CSVs only, not arbitrary malformed data or a broad CSV conformance suite. Ornith tmp/run.js was inspected but not separately executed." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/10886300e7.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/postmortem.md", "nativeTokensPerSec": 120.92, "nativeGeneratedTokens": 22332, "promptTokensPerSec": 3404.5, "usageSource": "engine-log", "durationMs": 287194, "toolCalls": 25, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.35906982421875 }, "peakRssMb": 10245, "peakGpuMemMb": 5801, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-data-wrangle", "model": "underdog-saluki-27b-iq2-mix", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-22T01:02:49.559Z", "finishedAt": "2026-10-22T01:47:10.458Z", "success": false, "failureClass": "model", "reason": "retry loop (long-path): sniff \"data-wrangle:7:targetnone:frxzh9f0:rp0:rf0\" stuck for 12m — team idle (4 new tool calls) with no sniff movement for 12m. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 5.3, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 5, "completeness": 5, "traceability": 5, "outputUsability": 4 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 unique customers and every field match the normalized source, but seven adjacent email pairs are unsorted. The readable selected cleaner produces the exact sorted reference in an isolated unchanged-code replay. Raw inputs/checker remain intact. This confirms a delivery/execution gap; replay output does not replace the final artifact.", "reviewInputBytes": 17760, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 review under the combined20KiB selected-text cap, including unchanged grounding and diagnostics.", "Exact inspected main helper replayed in a fresh copy with unchanged CSVs and existing empty output directory. Replay is derived diagnostic evidence, not a replacement scored output or official eval rerun.", "No original candidate edits, dependency installation, inference rerun or settings changes. Earlier helper body/reasoning excluded. Metadata does not establish every turn tool roster.", "Only supplied CSVs were exercised; no broad CSV parser conformance claim. Successful validate calls do not establish semantic completeness or sorting." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/5e24b4a2f0.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/postmortem.md", "nativeTokensPerSec": 12.52, "nativeGeneratedTokens": 21452, "promptTokensPerSec": 650.79, "usageSource": "engine-log", "durationMs": 2660899, "toolCalls": 22, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.30523681640625 }, "peakRssMb": 18431, "peakGpuMemMb": 10709, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-data-wrangle", "model": "qwen3.8-27b-q2", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 3, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-22T01:48:22.162Z", "finishedAt": "2026-10-22T02:38:09.235Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"data-wrangle:7:targetnone:fr1v2jaqq:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 5 tool calls (2 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 5.8, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 5, "completeness": 5, "traceability": 4, "outputUsability": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All 56 unique customers and every field match the normalized sources, but four adjacent email pairs are unsorted. The latest cleaner loses quoted-field boundaries and fails on Robert in an isolated unchanged-code replay, producing no output. Source CSVs and checker remain intact. The delivered data is complete; reproducibility and ordering are deficient.", "reviewInputBytes": 16444, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 review under the combined 20 KiB selected-text cap, including grounding and diagnostic excerpts.", "The exact inspected helper was replayed in a fresh directory with copied unchanged CSVs and an empty output directory; only supplied inputs were tested.", "Replay is diagnostic evidence, not a replacement artifact, original gate rerun or change to the frozen outcome. No candidate edits, dependencies or inference changes.", "Earlier helper source and reasoning excluded. Tool metadata does not establish every original turn tool roster. Successful validate tool status does not prove semantic sorting." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/b5aebdf59b.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/postmortem.md", "nativeTokensPerSec": 6.44, "nativeGeneratedTokens": 14808, "promptTokensPerSec": 460.57, "usageSource": "engine-log", "durationMs": 2987073, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3556976318359375 }, "peakRssMb": 18200, "peakGpuMemMb": 11687, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-data-wrangle", "model": "gemma4-12b-q4", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-22T02:39:11.993Z", "finishedAt": "2026-10-22T02:51:45.794Z", "success": false, "failureClass": "model", "reason": "retry loop (fast-path): sniff \"data-wrangle:4:targetnone:fr1jpp458:rp0:rf0\" stuck for 8m — 5 re-writes (11 tool calls) without sniff movement. Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 2.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 3, "completeness": 4, "traceability": 3, "outputUsability": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The 73 rows cover 54 of 56 customers, with 19 duplicate occurrences and 15 sorting inversions. Stale Zoe/Robert entries and a hybrid Omar id/date remain. The latest cleaner uses require in .mjs and fails before producing output in an isolated replay. Raw CSVs/checker are intact; the JSON is structured but needs substantial repair.", "reviewInputBytes": 20368, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-data-v1 assessment under the combined 20 KiB cap, including grounding and selected diagnostics.", "Exact inspected helper replayed in a fresh copied-input directory with existing empty out folder; no source repair, dependency change or original artifact overwrite. Only supplied CSVs tested.", "Earlier helper body and reasoning excluded; metadata does not establish every turn tool roster. Official outcome and fixed score are unchanged." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/42e133d613.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/postmortem.md", "nativeTokensPerSec": 132.41, "nativeGeneratedTokens": 44504, "promptTokensPerSec": 1631.78, "usageSource": "engine-log", "durationMs": 753801, "toolCalls": 30, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4302978515625 }, "peakRssMb": 19016, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-incident-postmortem", "model": "qwen3.5-9b-q4", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-22T02:52:57.315Z", "finishedAt": "2026-10-22T02:55:02.304Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 5, "clarity": 4 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Core cause, required sections and owned dated actions are present. It overstates degradation to 14:58, puts below-1% recovery at 14:53, invents a five-minute detection gap, and misstates 99% saturation onset. Recovery-table inconsistencies and unsupported review-bypass implications weaken an otherwise usable report.", "reviewInputBytes": 14828, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap, including grounding and selected diagnostics. Full report directly read; no rendered-document assessment.", "All five evidence inputs match the frozen scenario seeds. The timeline hotfix-start discrepancy is distinguished from model arithmetic errors and unsupported claims." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/1a66f886a3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/postmortem.md", "nativeTokensPerSec": 69.76, "nativeGeneratedTokens": 4308, "promptTokensPerSec": 2359.57, "usageSource": "engine-log", "durationMs": 124989, "toolCalls": 10, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3466796875 }, "peakRssMb": 9239, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-incident-postmortem", "model": "ornith1.5-9b-q4", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-22T02:55:56.158Z", "finishedAt": "2026-10-22T03:00:13.688Z", "success": false, "failureClass": "model", "reason": "repair-exhausted: bounded scenario feedback exhausted after 4 delivered-and-completed attempts on postmortem.md; missing signals stayed unchanged: grounded-core-facts. Last failure: grounded-core-facts: missing connection-pool saturation/exhaustion mechanism", "modelEligible": true, "fixedComposite": 6.4, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 4, "clarity": 4 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The mechanism and unknown impacts are explicit, with owned actions. It moves PR merge to rollout, asserts absent alerts, overextends 99% saturation, duplicates the cause paragraph and misaligns the impact table. The frozen mechanism check falsely rejects its hyphenated connection-pool wording; that finding does not excuse the factual defects.", "reviewInputBytes": 18071, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap, including grounding and selected diagnostics. Full report directly read; no rendered-document assessment.", "All five evidence inputs match the frozen scenario seeds. The timeline hotfix-start discrepancy is distinguished from model arithmetic errors and unsupported claims.", "Only the exact frozen mechanism predicate was probed; an in-memory compound-separator substitution changes its result. No full gate rerun, original artifact edit, reclassification or scoring override. Source identity matches the campaign manifest." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7e4feafcef.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/postmortem.md", "nativeTokensPerSec": 54.83, "nativeGeneratedTokens": 8961, "promptTokensPerSec": 2516.68, "usageSource": "engine-log", "durationMs": 257530, "toolCalls": 15, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4479522705078125 }, "peakRssMb": 10848, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-incident-postmortem", "model": "qwen3.5-4b-q4", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-22T03:01:07.864Z", "finishedAt": "2026-10-22T03:03:17.375Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.2, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 2, "functionalCompleteness": 5, "actionability": 3, "clarity": 4 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Complete sections and useful load-test/runbook actions are undermined by 10-second versus 8-second timeout, a one-minute claim for 14:46–14:53 deployment, invented 20-second recovery and unproven missing controls. The proposed timeout limit as 50% of pool capacity mixes incompatible units. Passing gates does not make these claims reliable.", "reviewInputBytes": 15297, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap, including grounding and selected diagnostics. Full report directly read; no rendered-document assessment.", "All five evidence inputs match the frozen scenario seeds. The timeline hotfix-start discrepancy is distinguished from model arithmetic errors and unsupported claims." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7dd90a1563.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/postmortem.md", "nativeTokensPerSec": 112.69, "nativeGeneratedTokens": 9026, "promptTokensPerSec": 3389.31, "usageSource": "engine-log", "durationMs": 129511, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2904815673828125 }, "peakRssMb": 7525, "peakGpuMemMb": 5801, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem", "model": "underdog-saluki-27b-iq2-mix", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-22T03:04:11.128Z", "finishedAt": "2026-10-22T03:29:25.248Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear, complete report with correct 21-minute degradation, primary-log hotfix timing, explicit unknown impacts and practical owned actions. It nevertheless asserts absent saturation alerts from silence and treats the canary load profile as established. Some source anchors and rounded saturation timing are imprecise.", "reviewInputBytes": 17515, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap, including grounding and selected diagnostics. Full report directly read; no rendered-document assessment.", "All five evidence inputs match the frozen scenario seeds. The timeline hotfix-start discrepancy is distinguished from model arithmetic errors and unsupported claims." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/836230ae45.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/postmortem.md", "nativeTokensPerSec": 12.35, "nativeGeneratedTokens": 13173, "promptTokensPerSec": 641.4, "usageSource": "engine-log", "durationMs": 1514120, "toolCalls": 9, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2955780029296875 }, "peakRssMb": 17082, "peakGpuMemMb": 10709, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-incident-postmortem", "model": "qwen3.8-27b-q2", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-22T03:30:31.716Z", "finishedAt": "2026-10-22T04:03:35.848Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Complete, clear report with the correct 21-minute degradation and primary hotfix timing. It puts the ten-minute monitoring window before mitigation rather than after, infers absent alerts from silence, and proposes a pool-sizing formula that multiplies by pool size again. Useful owned actions remain, but those claims need correction.", "reviewInputBytes": 16780, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "The source hotfix chronology discrepancy is distinguished from unsupported claims and arithmetic errors. Proposed deadlines and actions are not treated as existing commitments." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/739d1f5b43.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/postmortem.md", "nativeTokensPerSec": 5.45, "nativeGeneratedTokens": 10076, "promptTokensPerSec": 474.46, "usageSource": "engine-log", "durationMs": 1984132, "toolCalls": 6, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.323974609375 }, "peakRssMb": 14921, "peakGpuMemMb": 11685, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-incident-postmortem", "model": "gemma4-12b-q4", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-22T04:04:46.949Z", "finishedAt": "2026-10-22T04:08:00.739Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 4, "clarity": 3 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Required sections, the core mechanism and owned actions are present. The final analysis incorrectly calls 14:54 baseline recovery, overextends saturation, and overstates the effect of any latency. Repeated unheaded analysis and follow-up padding weaken clarity. The primary hotfix time is correct, but the report needs a concise factual reconciliation.", "reviewInputBytes": 15414, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "The source hotfix chronology discrepancy is distinguished from unsupported claims and arithmetic errors. Proposed deadlines and actions are not treated as existing commitments." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5dd5a06512.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/postmortem.md", "nativeTokensPerSec": 115.06, "nativeGeneratedTokens": 12104, "promptTokensPerSec": 1756.2, "usageSource": "engine-log", "durationMs": 193790, "toolCalls": 12, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4178009033203125 }, "peakRssMb": 18479, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-incident-postmortem", "model": "gemma4-e4b-q4", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-22T04:09:06.063Z", "finishedAt": "2026-10-22T04:10:51.711Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 2, "functionalCompleteness": 5, "actionability": 4, "clarity": 3 }, "advisoryMean": 3.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The cause, required sections and owned actions are present, but recovery timing contradicts itself: below-1% recovery is moved to 14:53, and restoration is called 15 minutes. Revert timing and the monitoring window are also wrong. Two repetitive appendices retain conflicting hotfix start times without explaining the source discrepancy.", "reviewInputBytes": 15083, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "The source hotfix chronology discrepancy is distinguished from unsupported claims and arithmetic errors. Proposed deadlines and actions are not treated as existing commitments." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/28624a466f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/postmortem.md", "nativeTokensPerSec": 184.13, "nativeGeneratedTokens": 10542, "promptTokensPerSec": 4231.7, "usageSource": "engine-log", "durationMs": 105648, "toolCalls": 12, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.315765380859375 }, "peakRssMb": 5817, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-ops-runbook-anomaly", "model": "ornith1.5-9b-q4", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-22T04:11:45.794Z", "finishedAt": "2026-10-22T04:13:33.998Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct values and an explicit fresh/under24 versus stale/71 halt, with no downstream files. Config and backup were read without an intervening log update, so the after-each-step rule was not fully followed. Several numbered citations are unresolved despite direct source paths. The final reports are useful and the stop was respected.", "reviewInputBytes": 6042, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6f1ce3c25c.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/postmortem.md", "nativeTokensPerSec": 55.96, "nativeGeneratedTokens": 3784, "promptTokensPerSec": 2607.94, "usageSource": "engine-log", "durationMs": 108204, "toolCalls": 17, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.331298828125 }, "peakRssMb": 7829, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-ops-runbook-anomaly", "model": "qwen3.5-4b-q4", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-22T04:14:27.350Z", "finishedAt": "2026-10-22T04:16:02.563Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct final STEP1–4 values and a concise halt report; later-step files are absent. Checksum, config and backup reads had no intervening log writes, so the requested incremental audit trail was not followed. Later rereads and rewrites complete the final log. An orphan reference is minor because the backup path is explicit.", "reviewInputBytes": 5701, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6c2d7e5b54.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/postmortem.md", "nativeTokensPerSec": 112.85, "nativeGeneratedTokens": 2359, "promptTokensPerSec": 3888.77, "usageSource": "engine-log", "durationMs": 95213, "toolCalls": 22, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.311065673828125 }, "peakRssMb": 11664, "peakGpuMemMb": 5805, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly", "model": "underdog-saluki-27b-iq2-mix", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-22T04:16:55.701Z", "finishedAt": "2026-10-22T04:20:28.465Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact ordered STEP1–4 values and a clear expected-versus-observed halt report. Log writes separate the successful checks, then a final rewrite follows the halt and reread. Both downstream files are absent. The concise final record meets the brief; intermediate write contents were not inspected.", "reviewInputBytes": 4036, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2a6bae2ce6.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/postmortem.md", "nativeTokensPerSec": 12.68, "nativeGeneratedTokens": 1525, "promptTokensPerSec": 633.62, "usageSource": "engine-log", "durationMs": 212764, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2573089599609375 }, "peakRssMb": 14457, "peakGpuMemMb": 10707, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-ops-runbook-anomaly", "model": "qwen3.8-27b-q2", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-22T04:21:22.504Z", "finishedAt": "2026-10-22T04:27:05.440Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact STEP1–4 values, a clear fresh/under24 versus stale/71 comparison, and explicit skipped downstream steps. Source reads and intervening log updates preserve the successful-step sequence; the final rewrite follows the halt. Neither downstream file exists. Complete and usable, without exceptional polish.", "reviewInputBytes": 4158, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d8d474133f.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/postmortem.md", "nativeTokensPerSec": 6.34, "nativeGeneratedTokens": 1412, "promptTokensPerSec": 457.58, "usageSource": "engine-log", "durationMs": 342936, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.25994873046875 }, "peakRssMb": 16659, "peakGpuMemMb": 11683, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-ops-runbook-anomaly", "model": "gemma4-12b-q4", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-22T04:28:07.687Z", "finishedAt": "2026-10-22T04:29:51.677Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct STEP1–4 values and clear halt conditions. Writes separate the successful checks, and the final log includes stale/71; later-step files are absent. Two failed artifact writes precede recovery and further report revisions. The final records are concise and usable; prior draft contents were not inspected.", "reviewInputBytes": 5058, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/0d1c8bbd69.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/postmortem.md", "nativeTokensPerSec": 121.87, "nativeGeneratedTokens": 1803, "promptTokensPerSec": 1819.99, "usageSource": "engine-log", "durationMs": 103990, "toolCalls": 22, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3180999755859375 }, "peakRssMb": 18372, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-ops-runbook-anomaly", "model": "gemma4-e4b-q4", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-22T04:31:29.072Z", "finishedAt": "2026-10-22T04:32:41.236Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct ordered verification values and a grounded halt report, with no downstream outputs. Log writes separate the successful checks; later replacements finalize the reports. A stray citation label is minor because the backup source and failed conditions are explicit. The final record meets the brief.", "reviewInputBytes": 4369, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5cefa7cb82.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/postmortem.md", "nativeTokensPerSec": 195.71, "nativeGeneratedTokens": 3906, "promptTokensPerSec": 4571.88, "usageSource": "engine-log", "durationMs": 72164, "toolCalls": 15, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2905120849609375 }, "peakRssMb": 5779, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-ops-runbook-anomaly", "model": "qwen3.5-9b-q4", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-22T04:33:34.811Z", "finishedAt": "2026-10-22T04:34:43.407Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Exact STEP1–4 values, concise expected-versus-observed halt details, and no downstream files. Source checks alternate with log writes, followed by the halt and a final log rewrite after rereading the instructions. The final documents are complete and easy to use; intermediate draft contents were not inspected.", "reviewInputBytes": 3812, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Compact tool metadata establishes operation order but not the contents of earlier writes. No forbidden downstream files were found. No official gate rerun." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b531fdf7c3.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/postmortem.md", "nativeTokensPerSec": 71.68, "nativeGeneratedTokens": 1501, "promptTokensPerSec": 2416.23, "usageSource": "engine-log", "durationMs": 68596, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2723846435546875 }, "peakRssMb": 9638, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-plan-and-estimate", "model": "qwen3.5-4b-q4", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-22T04:35:37.033Z", "finishedAt": "2026-10-22T04:36:39.409Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 3, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All workstreams, internal owners, physical move and workshop completion are covered. However, move execution omits network and packing prerequisites, and print/scan setup bypasses provisioning. The mid-August booking target is not reconciled with the August25 sign-off assumption. Clear structure masks material scheduling gaps.", "reviewInputBytes": 4977, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Dependency and claim review plus the existing pure plan check; gate acceptance does not establish execution feasibility. No candidate edits or replacement official result." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c0534ce639.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/postmortem.md", "nativeTokensPerSec": 101.3, "nativeGeneratedTokens": 2534, "promptTokensPerSec": 3900.33, "usageSource": "engine-log", "durationMs": 62376, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.299072265625 }, "peakRssMb": 7239, "peakGpuMemMb": 5801, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate", "model": "underdog-saluki-27b-iq2-mix", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-22T04:37:33.111Z", "finishedAt": "2026-10-22T04:40:04.877Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Covers internal ownership, approved layout before booking, workshop acceptance, packing and the physical move. The network test task is present, but move execution depends on cabling T5 rather than completed testing T6; handover also omits T6. The prose pre-day-one constraint therefore lacks an explicit dependency. Otherwise concise and useful.", "reviewInputBytes": 4365, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Dependency and claim review plus the existing pure plan check; gate acceptance does not establish execution feasibility. No candidate edits or replacement official result." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/bb5cc90e34.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/postmortem.md", "nativeTokensPerSec": 12.99, "nativeGeneratedTokens": 1333, "promptTokensPerSec": 694.71, "usageSource": "engine-log", "durationMs": 151766, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.23468017578125 }, "peakRssMb": 9878, "peakGpuMemMb": 10707, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-plan-and-estimate", "model": "qwen3.8-27b-q2", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-22T04:40:58.675Z", "finishedAt": "2026-10-22T04:47:52.933Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Complete internal ownership, layout approval before booking, workshop build, network/print tests and the actual move. Move execution depends on network readiness but omits its packing/schedule task T10. A studio speed threshold is undefined, and the sign-off deadline is left unspecified. The plan is clear and largely executable.", "reviewInputBytes": 5804, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Dependency and claim review plus the existing pure plan check; gate acceptance does not establish execution feasibility. No candidate edits or replacement official result." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/68a57cd9ee.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/postmortem.md", "nativeTokensPerSec": 4.19, "nativeGeneratedTokens": 1451, "promptTokensPerSec": 510.19, "usageSource": "engine-log", "durationMs": 414258, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.235748291015625 }, "peakRssMb": 11764, "peakGpuMemMb": 11683, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-plan-and-estimate", "model": "gemma4-12b-q4", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-22T04:48:51.066Z", "finishedAt": "2026-10-22T04:49:27.783Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 3, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear internal owners, signed layout before booking, and a dated network installation/test task. The plan stops at booking movers and confirming the workshop start date: no task executes the physical move or accepts the completed workshop. The risks mention completion, but those actions have no corresponding delivery tasks.", "reviewInputBytes": 2999, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB input cap. Sources are grounding, not generated deliverables.", "Complete selected outputs were read directly; no rendered-document assessment.", "Dependency and claim review plus the existing pure plan check; gate acceptance does not establish execution feasibility. No candidate edits or replacement official result." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2051814134.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/postmortem.md", "nativeTokensPerSec": 108.39, "nativeGeneratedTokens": 1140, "promptTokensPerSec": 1820.92, "usageSource": "engine-log", "durationMs": 36717, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.262298583984375 }, "peakRssMb": 11319, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-plan-and-estimate", "model": "gemma4-e4b-q4", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-22T04:50:27.159Z", "finishedAt": "2026-10-22T04:51:00.478Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct internal owners, signed layout before booking, completed workshop, network tests and an actual move depending on readiness. Workshop execution depends on its design but not final integrated floor-plan sign-off. Milestone dates are unspecified despite a proposed buffer week. Otherwise clear, complete and usable.", "reviewInputBytes": 3948, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap. Source files are grounding, not generated artifacts.", "Complete plan read directly; no rendered-document assessment. Dependency and source-fidelity review is separate from the existing pure plan check.", "No candidate edits, gate changes or replacement official results. Effort estimates are not expressly required by the brief." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/771632c70d.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/postmortem.md", "nativeTokensPerSec": 175.88, "nativeGeneratedTokens": 2640, "promptTokensPerSec": 4431.42, "usageSource": "engine-log", "durationMs": 33319, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.24981689453125 }, "peakRssMb": 4538, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-plan-and-estimate", "model": "qwen3.5-9b-q4", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-22T04:51:58.531Z", "finishedAt": "2026-10-22T04:52:44.569Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 3, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The plan covers all workstreams with internal owners and final readiness checks. It invents an approved status and names workshop contractor Brightline as the movers. Booking by September10 is not reconciled with a September15 sign-off risk threshold. Network installation acceptance checks a diagram, and moving lacks an explicit network prerequisite.", "reviewInputBytes": 4250, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap. Source files are grounding, not generated artifacts.", "Complete plan read directly; no rendered-document assessment. Dependency and source-fidelity review is separate from the existing pure plan check.", "No candidate edits, gate changes or replacement official results. Effort estimates are not expressly required by the brief." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/05576a6a76.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/postmortem.md", "nativeTokensPerSec": 51.99, "nativeGeneratedTokens": 1077, "promptTokensPerSec": 2531.24, "usageSource": "engine-log", "durationMs": 46038, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2496185302734375 }, "peakRssMb": 7363, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-plan-and-estimate", "model": "ornith1.5-9b-q4", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-22T04:53:44.268Z", "finishedAt": "2026-10-22T04:54:41.905Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear internal owners, signed layout before booking, workshop completion and network checks. No task executes the physical move. Final readiness omits the workshop build dependency, and print/scan setup bypasses broadband readiness. Risk prose mixes moving-day fault discovery with testing before packing. The plan is readable but lacks a complete execution chain.", "reviewInputBytes": 3550, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined 20 KiB cap. Source files are grounding, not generated artifacts.", "Complete plan read directly; no rendered-document assessment. Dependency and source-fidelity review is separate from the existing pure plan check.", "No candidate edits, gate changes or replacement official results. Effort estimates are not expressly required by the brief." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b3bfdf6244.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/postmortem.md", "nativeTokensPerSec": 58.07, "nativeGeneratedTokens": 1794, "promptTokensPerSec": 2756.19, "usageSource": "engine-log", "durationMs": 57637, "toolCalls": 4, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2450408935546875 }, "peakRssMb": 6709, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis", "model": "underdog-saluki-27b-iq2-mix", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "underdog-saluki-27b-iq2-mix" }, "startedAt": "2026-10-22T04:55:40.510Z", "finishedAt": "2026-10-22T05:00:22.033Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winning date, budget, allocation and ownership, with all three source-bound resolutions and historical values quarantined. Feature scope is preserved. Open questions sensibly identify missing migration timing, contractor roles, scope detail and current meeting cadence. Clear, complete synthesis that meets the brief.", "reviewInputBytes": 3642, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2a73de31f0.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/postmortem.md", "nativeTokensPerSec": 12.4, "nativeGeneratedTokens": 2668, "promptTokensPerSec": 646.87, "usageSource": "engine-log", "durationMs": 281523, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2657623291015625 }, "peakRssMb": 10372, "peakGpuMemMb": 10707, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.8-27b-q2-core-conflict-synthesis", "model": "qwen3.8-27b-q2", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.8-27b-q2" }, "startedAt": "2026-10-22T05:01:16.008Z", "finishedAt": "2026-10-22T05:24:46.811Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winners, allocation and explicit source-priority reconciliation. It carries historical Tuesday meetings into the plan as the only unsuperseded element, although current validity is unknown; the open questions acknowledge this. Migration-start and budget-history questions are useful. Clear overall, but an unverified cadence remains an adopted plan item.", "reviewInputBytes": 4142, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/f8c042a277.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/postmortem.md", "nativeTokensPerSec": 6.03, "nativeGeneratedTokens": 7368, "promptTokensPerSec": 473.38, "usageSource": "engine-log", "durationMs": 1410803, "toolCalls": 26, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.306365966796875 }, "peakRssMb": 17871, "peakGpuMemMb": 11685, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-12b-q4-core-conflict-synthesis", "model": "gemma4-12b-q4", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-12b-q4" }, "startedAt": "2026-10-22T05:25:56.782Z", "finishedAt": "2026-10-22T05:27:41.968Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 3, "clarity": 3 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The winning values, budget breakdown and three reconciliations are correct. A duplicated Summary/Method/Findings/Recommendations block precedes the requested brief, weakening clarity. Current campaign/engineering leads are omitted, and Open questions says none despite unconfirmed historical cadence and missing migration timing. Useful facts, incomplete synthesis.", "reviewInputBytes": 2726, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7711e0abf1.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/postmortem.md", "nativeTokensPerSec": 129.96, "nativeGeneratedTokens": 3468, "promptTokensPerSec": 1933.96, "usageSource": "engine-log", "durationMs": 105186, "toolCalls": 14, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.282958984375 }, "peakRssMb": 16906, "peakGpuMemMb": 8701, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-gemma4-e4b-q4-core-conflict-synthesis", "model": "gemma4-e4b-q4", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "gemma4-e4b-q4" }, "startedAt": "2026-10-22T05:28:46.207Z", "finishedAt": "2026-10-22T05:30:26.119Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 3, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Concise, accurate winning date, budget, ownership and scope, with all three explicit resolutions and source filenames. The synthesis omits the engineering lead, budget allocation and six-week migration detail. It identifies no open questions despite missing timing details and uncertain historical meeting cadence. Clear but less complete and actionable than the source pack allows.", "reviewInputBytes": 2286, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/9e719b1005.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/postmortem.md", "nativeTokensPerSec": 195.06, "nativeGeneratedTokens": 8607, "promptTokensPerSec": 4671.54, "usageSource": "engine-log", "durationMs": 99912, "toolCalls": 18, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2962493896484375 }, "peakRssMb": 6482, "peakGpuMemMb": 4139, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-9b-q4-core-conflict-synthesis", "model": "qwen3.5-9b-q4", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-9b-q4" }, "startedAt": "2026-10-22T05:31:20.018Z", "finishedAt": "2026-10-22T05:33:41.842Z", "success": false, "failureClass": "model", "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 0 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", "modelEligible": true, "fixedComposite": 6, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 4, "actionability": 4, "clarity": 4 }, "advisoryMean": 4, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winning values, allocation and three reconciliations, but the required backticked filenames are replaced with numbered references and bold names. Those numbers have no matching numbered source definitions. It also calls the contractor allocation approved without evidence. The useful final brief does not satisfy the explicit citation contract.", "reviewInputBytes": 3380, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4bc61b4479.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/postmortem.md", "nativeTokensPerSec": 83.69, "nativeGeneratedTokens": 4407, "promptTokensPerSec": 2579.95, "usageSource": "engine-log", "durationMs": 141824, "toolCalls": 27, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.3379364013671875 }, "peakRssMb": 14142, "peakGpuMemMb": 8181, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-ornith1.5-9b-q4-core-conflict-synthesis", "model": "ornith1.5-9b-q4", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "ornith1.5-9b-q4" }, "startedAt": "2026-10-22T05:34:34.789Z", "finishedAt": "2026-10-22T05:39:05.958Z", "success": false, "failureClass": "model", "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 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", "modelEligible": true, "fixedComposite": 5.3, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 3, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.25, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winning values, clear source-priority resolutions and useful unknowns. However, the superseded240,000 budget reappears in Open questions, violating explicit quarantine. The30,000 difference is valid arithmetic, despite the gate calling it unsupported. Historical cadence is now kept uncertain. The main remaining defect is contract compliance.", "reviewInputBytes": 3634, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/615f79172a.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/postmortem.md", "nativeTokensPerSec": 58.49, "nativeGeneratedTokens": 9906, "promptTokensPerSec": 2648.38, "usageSource": "engine-log", "durationMs": 271169, "toolCalls": 49, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.4467926025390625 }, "peakRssMb": 14699, "peakGpuMemMb": 7335, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r3-qwen3.5-4b-q4-core-conflict-synthesis", "model": "qwen3.5-4b-q4", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o", "journal": "B", "journalRoot": "evals/runs/small-model-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 64K", "contextTokens": 65536, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 65536, "totalTokens": 65536, "slots": 1, "kvCacheType": "f16", "launchModel": "qwen3.5-4b-q4" }, "startedAt": "2026-10-22T05:40:00.221Z", "finishedAt": "2026-10-22T05:40:57.306Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 5, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct winners, allocation, source-bound conflict table and historical-value quarantine. Ownership and scope are clear, with useful questions about cadence and contractor roles. Calling September1 the only viable launch window exceeds the engineering memo, which chooses that date without ruling out every later option. Otherwise concise and usable.", "reviewInputBytes": 3110, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded frozen campaign-doc-v1 assessment under the combined20KiB input cap. Grounding files are not generated outputs.", "Full document directly read; no rendered-document assessment. Source conflicts, supersession, arithmetic and unknowns reviewed manually.", "Original outcomes, classifications and fixed scores are preserved. No checker rerun or candidate edits." ], "runDir": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o", "judgmentPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/codex-inline-judge.json", "evidencePath": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5bbfc53303.json", "postmortemPath": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/postmortem.md", "nativeTokensPerSec": 108.6, "nativeGeneratedTokens": 2895, "promptTokensPerSec": 3626.51, "usageSource": "engine-log", "durationMs": 57085, "toolCalls": 7, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.2912750244140625 }, "peakRssMb": 5939, "peakGpuMemMb": 5801, "runId": "2026-10-20-rtx5070ti-core-continuation-64k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-tictactoe", "model": "nanbeige4.2-3b-q4", "scenario": "tictactoe", "repetition": 1, "trialId": "tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby", "journal": "N1", "journalRoot": "evals/runs/nanbeige-32k-core-2026-10-20", "runtimeCohort": "Corrected runtime / 32K initial Nanbeige", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-20T16:55:19.364Z", "finishedAt": "2026-10-20T16:59:13.820Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.8, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 5, "codeQuality": 5, "polish": 4 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Win, draw, repeat-click guards, post-win guards and complete reset all work without browser errors. The initial turn indicator is blank until a move/reset; cell buttons have no accessible names. Fixed-width cells spread across wide grid tracks make the board look disconnected and off-center. Mobile fits, but styling and feedback are basic.", "reviewInputBytes": 3750, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Reviewed the saved final artifact, with external HTTP(S) requests blocked.", "Visual inspection covered desktop and 390px mobile layouts; no exhaustive accessibility audit.", "This judgment applies only to the new 32K cohort. It is separate from the deterministic pass and measured generation throughput." ], "runDir": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby", "judgmentPath": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/9cb7a25c13.json", "postmortemPath": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/postmortem.md", "nativeTokensPerSec": 44.29, "nativeGeneratedTokens": 7566, "promptTokensPerSec": 1548.52, "usageSource": "engine-log", "durationMs": 234456, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.55450439453125 }, "peakRssMb": 10353, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-petshop", "model": "nanbeige4.2-3b-q4", "scenario": "petshop", "repetition": 1, "trialId": "petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4", "journal": "N1", "journalRoot": "evals/runs/nanbeige-32k-core-2026-10-20", "runtimeCohort": "Corrected runtime / 32K initial Nanbeige", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-20T17:02:05.216Z", "finishedAt": "2026-10-20T17:27:03.417Z", "success": false, "failureClass": "model", "reason": "repair-aborted: kazuki/7469f8a8 exhausted its bounded automatic recovery allowance at the current checked progress; last error: [llama-cpp] /v1/chat/completions returned 400 Bad Request: {\"error\":{\"code\":400,\"message\":\"Failed to initialize samplers: Unexpected empty grammar stack after accepting piece: assistant (13886)\",\"type\":\"invalid_request_error\"}}", "modelEligible": true, "fixedComposite": 1.7, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 2, "codeQuality": 3, "polish": 2 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive cards and section navigation work, but the required PNG logo is missing and the SVG is unreferenced. Stats initialization and Home clicks throw on missing element IDs; displayed totals contradict the pet list. Add to Cart has no handler. Generic copy and no shop contact details leave a partial storefront. External placeholder imagery was not assessed.", "reviewInputBytes": 10252, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Offline saved artifact review; HTTP(S) blocked. The six Picsum image URLs were not assessed online.", "The local logo.png reference is missing regardless of network access; the saved SVG is not used by the page.", "No generate_image call appears in trial facts. Artifact review does not override the original terminal grammar failure or deterministic score.", "Targeted click checks only; no artifact repairs." ], "runDir": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4", "judgmentPath": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/3aeae200b9.json", "postmortemPath": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/postmortem.md", "nativeTokensPerSec": 48.42, "nativeGeneratedTokens": 42924, "promptTokensPerSec": 1481.48, "usageSource": "engine-log", "durationMs": 1498201, "toolCalls": 66, "redFlags": [ { "gezel": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/sessions/kenji", "tool": "(no write_file this turn)", "argsSummary": "content_chars=1969 h2_count=2 preview=\"## 🎯 My Judgment Call — Proceeding Now Since `generate_image` isn't available and you asked me to decide, here's my ca…\"", "atTurn": 8, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." } ], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 6, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.70794677734375 }, "peakRssMb": 14782, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-tankcombat", "model": "nanbeige4.2-3b-q4", "scenario": "tankcombat", "repetition": 1, "trialId": "tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf", "journal": "N1", "journalRoot": "evals/runs/nanbeige-32k-core-2026-10-20", "runtimeCohort": "Corrected runtime / 32K initial Nanbeige", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-20T17:28:18.652Z", "finishedAt": "2026-10-20T17:59:16.227Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,787 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 3.3, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 0, "codeQuality": 2, "polish": 1 }, "advisoryMean": 1.5, "advisoryEligible": false, "advisoryExclusion": "Recorded context-overflow infrastructure failure; partial artifact reviewed for diagnostics, excluded from model advisory aggregates.", "justification": "First render throws 'Player is not defined', leaving a static grid and overlapping enemies. ArrowRight, D and Space produce no movement or firing. Initialization clamps entities using zero dimensions; gameLoop never schedules another update. Projectile ownership/removal logic is also broken. Fixed canvas clips badly at 390px. No playable game.", "reviewInputBytes": 9425, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The trial ended on an infrastructure-classified context overflow. This scores only the captured partial artifact and is excluded from model advisory aggregates.", "The fatal Player reference and lack of input effects were replayed. Further initialization, loop and projectile defects are source findings; repaired gameplay was not tested.", "Desktop Chromium at two viewport widths; not mobile device emulation." ], "runDir": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf", "judgmentPath": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/13d0445595.json", "postmortemPath": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/postmortem.md", "nativeTokensPerSec": 44.32, "nativeGeneratedTokens": 40711, "promptTokensPerSec": 1297.46, "usageSource": "engine-log", "durationMs": 1857575, "toolCalls": 107, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 4, "forceFit": 0, "compactStarts": 2, "compactFailed": 2, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.74224853515625 }, "peakRssMb": 11730, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-schema-migration", "model": "nanbeige4.2-3b-q4", "scenario": "schema-migration", "repetition": 1, "trialId": "schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w", "journal": "N2", "journalRoot": "evals/runs/nanbeige-32k-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 32K Nanbeige continuation", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-20T18:15:01.863Z", "finishedAt": "2026-10-20T18:17:36.034Z", "success": false, "failureClass": "model", "reason": "repair-aborted: riley/7e093873 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.", "modelEligible": true, "fixedComposite": 3.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 4, "robustness": 2, "scopeDiscipline": 3 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Types/store/handlers use split names, but migration erases one-word names and drops extra fields despite its preservation claim. Tests and MIGRATION.md are missing. CreateUserInput now wrongly requires id although the store generates it; rendered cards lose their closing div and log separators change. Production source compiles, but the normal create-input consumer fails typing.", "reviewInputBytes": 3201, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Selected changed/generated textual files are within 20 KiB; unchanged seed files were excluded from the artifact assessment.", "Production exports and a virtual TypeScript consumer were probed. Model-authored tests were inspected, not executed.", "Targeted cases are not exhaustive." ], "runDir": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/advisory-code-evidence/d86e968253.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/postmortem.md", "nativeTokensPerSec": 58.83, "nativeGeneratedTokens": 3682, "promptTokensPerSec": 2194.77, "usageSource": "engine-log", "durationMs": 154171, "toolCalls": 40, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 7, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.666748046875 }, "peakRssMb": 9349, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-failing-tests-spec", "model": "nanbeige4.2-3b-q4", "scenario": "failing-tests-spec", "repetition": 1, "trialId": "failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy", "journal": "N2", "journalRoot": "evals/runs/nanbeige-32k-core-continuation-2026-10-20", "runtimeCohort": "Corrected runtime / 32K Nanbeige continuation", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-20T18:18:38.906Z", "finishedAt": "2026-10-20T18:22:32.595Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,806 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 3.3, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 3, "codeQuality": 3, "robustness": 2, "scopeDiscipline": 5 }, "advisoryMean": 3.25, "advisoryEligible": false, "advisoryExclusion": "Recorded context-overflow infrastructure failure; captured partial implementation is diagnostic only and excluded from model advisory aggregates.", "justification": "Basic order flow, history and unknown-event rejection work, but shipped orders can be cancelled, the expedited option is ignored, and expedite stays paid instead of shipping. Strict compilation fails on string indexing and assignment to readonly state. Seed tests/configs are untouched. Advisory probes pass80/84 ordinary and36/36 special-name cases; partial output remains infra-excluded.", "reviewInputBytes": 1631, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Trial terminated on context overflow (32806 requested versus32768), not a completed capability verdict.", "Captured implementation was reviewed and probed; its advisory score is excluded from model aggregates.", "Targeted public-API cases and TypeScript diagnostics are advisory; unchanged seeded tests were not treated as generated output." ], "runDir": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/advisory-code-evidence/bbdae3e60a.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/postmortem.md", "nativeTokensPerSec": 43.85, "nativeGeneratedTokens": 4912, "promptTokensPerSec": 1108.81, "usageSource": "engine-log", "durationMs": 233689, "toolCalls": 49, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 3, "forceFit": 0, "compactStarts": 2, "compactFailed": 2, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.7513427734375 }, "peakRssMb": 10787, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-symptom-debug", "model": "nanbeige4.2-3b-q4", "scenario": "symptom-debug", "repetition": 1, "trialId": "symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q", "journal": "N3", "journalRoot": "evals/runs/nanbeige-32k-core-continuation-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige remaining cells (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T05:57:06.417Z", "finishedAt": "2026-10-22T06:06:07.340Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line correction removes the cursor skip and passes all 62 targeted checks without input mutation. The adjacent comment still incorrectly describes resuming one past the cursor. Acceptance script and package are unchanged; the fix stays within scope.", "reviewInputBytes": 16654, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The full selected textual input including grounding and diagnostics is 16654 bytes, below the combined 20 KiB cap.", "The pure production module was inspected before importing its exports. Sixty-two varied-size walk, invalid-input and direct-cursor probes were run with ordinary arrays/options; the frozen official gate was not rerun or modified.", "Acceptance and package files remain byte-identical to their seeds. No additional candidate files were edited or executed.", "Targeted cases are not exhaustive; the original decoder behavior outside the requested repair is preserved." ], "runDir": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-code-evidence/f2a64617c8.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/postmortem.md", "nativeTokensPerSec": 5.76, "nativeGeneratedTokens": 2383, "promptTokensPerSec": 232.97, "usageSource": "engine-log", "durationMs": 540923, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5140380859375 }, "peakRssMb": 3882, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-data-wrangle", "model": "nanbeige4.2-3b-q4", "scenario": "data-wrangle", "repetition": 1, "trialId": "data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd", "journal": "N3", "journalRoot": "evals/runs/nanbeige-32k-core-continuation-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige remaining cells (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T06:07:16.311Z", "finishedAt": "2026-10-22T07:09:44.812Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,890 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole", "modelEligible": false, "fixedComposite": 2, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-data-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "Infrastructure exclusion", "justification": "Required out/customers.json is absent. A cleaning helper exists but fails syntax-only validation on an unterminated regular-expression group. Its installed-script invocation failed; repeated source reads preceded the preserved 32890/32768-token context overflow. No data-quality scores are assigned to an absent deliverable.", "reviewInputBytes": 11984, "protocolDeviation": false, "reviewLimitations": [ "No requested customer JSON exists, so no advisory axes or mean are assigned. This remains an infrastructure exclusion.", "The full helper was inspected; node --check only parses it and confirms a syntax error. Candidate code was not executed, repaired or used to fabricate output.", "Raw CSVs and the checker are byte-identical to frozen seeds. Full logs/recording are hashed for provenance; only compact selected metadata and errors enter the review.", "Model identity was known; review is unblinded.", "Model identity was known during review." ], "runDir": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-data-evidence/d24aa7033c.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/postmortem.md", "nativeTokensPerSec": 2.99, "nativeGeneratedTokens": 9371, "promptTokensPerSec": 94.12, "usageSource": "engine-log", "durationMs": 3748501, "toolCalls": 34, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 1, "loopHalts": 0, "maxContextFill": 0.693603515625 }, "peakRssMb": 3845, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-incident-postmortem", "model": "nanbeige4.2-3b-q4", "scenario": "incident-postmortem", "repetition": 1, "trialId": "incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y", "journal": "N4", "journalRoot": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after third context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T08:02:02.674Z", "finishedAt": "2026-10-22T08:29:07.156Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 33,026 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 2.5, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-doc-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "Infrastructure exclusion", "justification": "No postmortem.md or other authored deliverable exists. The five evidence files are unchanged. Facts record 29 read-only tool calls before the preserved 33026/32768-token context overflow; forced compaction removed no messages. No advisory axes or mean are assigned to missing output.", "reviewInputBytes": 5754, "protocolDeviation": false, "reviewLimitations": [ "No requested document exists; the seeded evidence pack is not scored as model output.", "Original infrastructure classification remains unchanged. No candidate content was created, repaired or executed.", "Selected diagnostics remain below20KiB. Full source logs are provenance only. Model identity was known.", "Model identity was known during review." ], "runDir": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/advisory-document-evidence/1f6ebd7088.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/postmortem.md", "nativeTokensPerSec": 2.95, "nativeGeneratedTokens": 3043, "promptTokensPerSec": 91.61, "usageSource": "engine-log", "durationMs": 1624482, "toolCalls": 29, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 1, "loopHalts": 0, "maxContextFill": 0.74896240234375 }, "peakRssMb": 4049, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly", "model": "nanbeige4.2-3b-q4", "scenario": "ops-runbook-anomaly", "repetition": 1, "trialId": "ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e", "journal": "N5", "journalRoot": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fourth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T09:03:37.121Z", "finishedAt": "2026-10-22T09:47:44.492Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 8.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Both reports give the correct checks, stale/71-hour backup anomaly and halt; no later-step outputs exist. The trace reads checksum, config and backup before the next log write, so the mandated per-step recording order is incomplete. Clear source references and restart conditions make the final documents usable.", "reviewInputBytes": 13846, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "All selected text, including grounding and diagnostics, counts toward the combined20KiB cap.", "Tool metadata proves operation order but not the contents of earlier writes. No earlier draft was reconstructed or scored.", "Final prose and source fidelity were read directly; no rendered-document review or official checker rerun.", "The failed-write diagnostics suggest a quotation/evidence-matching issue, but no isolated predicate reproduction was performed.", "Performance under the observed33W/810MHz hardware restriction is separate from artifact quality and fixed scores." ], "runDir": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/0ef033ed95.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/postmortem.md", "nativeTokensPerSec": 3.38, "nativeGeneratedTokens": 7229, "promptTokensPerSec": 163.6, "usageSource": "engine-log", "durationMs": 2647371, "toolCalls": 32, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 3, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.648284912109375 }, "peakRssMb": 9118, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-plan-and-estimate", "model": "nanbeige4.2-3b-q4", "scenario": "plan-and-estimate", "repetition": 1, "trialId": "plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld", "journal": "N5", "journalRoot": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fourth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T09:49:18.717Z", "finishedAt": "2026-10-22T10:01:10.427Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear internal owners, signed layout before mover booking, operational network checks and an actual move. Workshop work ends at briefing/scheduling, with no owned build acceptance; readiness omits that branch. T6 requires occupied rooms before the move in T8. The plan is readable but its completion chain is incomplete.", "reviewInputBytes": 7589, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Direct prose, owner, dependency and done-state review; no rendered-document inspection or real-world feasibility test.", "The two source files and exact task brief provide grounding; only plan.md is scored as model output.", "No official checker rerun, candidate edit or automatic score inference.", "Runtime/context and the observed Oct22 GPU power restriction are separate performance boundaries." ], "runDir": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/7e3af2da10.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/postmortem.md", "nativeTokensPerSec": 3.27, "nativeGeneratedTokens": 1834, "promptTokensPerSec": 190.9, "usageSource": "engine-log", "durationMs": 711710, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5179443359375 }, "peakRssMb": 3887, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r1-nanbeige4.2-3b-q4-core-conflict-synthesis", "model": "nanbeige4.2-3b-q4", "scenario": "conflict-synthesis", "repetition": 1, "trialId": "conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet", "journal": "N5", "journalRoot": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fourth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T10:02:22.477Z", "finishedAt": "2026-10-22T11:03:24.750Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"conflict-synthesis:8:targetnone:fropcjni:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 25 tool calls (0 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 4.8, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 4, "clarity": 4 }, "advisoryMean": 3.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct date, budget allocation and DRI, with three reconciliations. Superseded budget and Priya reappear in Open questions, violating quarantine; citations mostly use numbers/bold names instead of required backticked filenames. It wrongly declares Tuesday syncs superseded. The30,000 difference is valid arithmetic, not an invented source budget.", "reviewInputBytes": 14101, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "The entire final synthesis, all six unchanged source files, exact brief and compact actual tool/fact diagnostics are included below20KiB. Only the synthesis is model output.", "Direct prose/source/citation review; no official checker rerun or rendered-document inspection.", "The30,000 difference is a correct derived arithmetic value. Existing numerical gate objections and all official classifications remain unchanged." ], "runDir": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/ecc910ec66.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/postmortem.md", "nativeTokensPerSec": 3.18, "nativeGeneratedTokens": 8465, "promptTokensPerSec": 124.4, "usageSource": "engine-log", "durationMs": 3662273, "toolCalls": 39, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.562835693359375 }, "peakRssMb": 11224, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-tictactoe", "model": "nanbeige4.2-3b-q4", "scenario": "tictactoe", "repetition": 2, "trialId": "tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4", "journal": "N5", "journalRoot": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fourth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T11:05:01.738Z", "finishedAt": "2026-10-22T11:49:29.773Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 8.3, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 4, "codeQuality": 4, "polish": 2 }, "advisoryMean": 3, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Turns, occupied-cell protection, row win, post-win lock, draw and keyboard input work. Empty cells are only6px high; rows expand as marks appear. CSS expects uppercase X/O but JavaScript adds lowercase classes, so colors fail. No restart control, viewport meta, empty-cell labels or live status. Core logic works, but the initial board is hard to use.", "reviewInputBytes": 11337, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full source, exact brief, selected actual tool/fact diagnostics and compact replay observations are included below20KiB.", "Desktop Chromium replay at1200x900 and390x844 with HTTP(S) blocked; narrow desktop view is not mobile-device emulation.", "Screenshots at empty, winner and narrow states were viewed. Row-win, draw, duplicate click, post-win and Space-key checks are targeted, not exhaustive game-tree testing.", "No in-page restart exists; replay resets used browser reload. No source edits or official gate rerun." ], "runDir": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/postmortem.md", "nativeTokensPerSec": 3.91, "nativeGeneratedTokens": 6503, "promptTokensPerSec": 186.14, "usageSource": "engine-log", "durationMs": 2668035, "toolCalls": 17, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.586395263671875 }, "peakRssMb": 9906, "peakGpuMemMb": 8219, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-petshop", "model": "nanbeige4.2-3b-q4", "scenario": "petshop", "repetition": 2, "trialId": "petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x", "journal": "N5", "journalRoot": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fourth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T11:51:26.327Z", "finishedAt": "2026-10-22T12:32:06.163Z", "success": false, "failureClass": "model", "reason": "retry loop (stall-path): sniff \"pet-shop-website/workspace/index.html:3:targetnone:frnone:rp0:rf0\" stuck for 18m — artifact scored but stalled 18m despite 0 tool calls (0 re-writes). Artifact produced but never reached success.", "modelEligible": true, "fixedComposite": 4.7, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 4, "functionalCompleteness": 2, "codeQuality": 3, "polish": 2 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Responsive cards are readable, but the required generated PNG is missing. Add and View do nothing: selection changes only when a button already has the selected class, which no initial Add has. The summary stays hidden and prices are never rendered. No browser exceptions, but the page promises interactions it cannot deliver.", "reviewInputBytes": 13151, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Offline desktop Chromium at1200x900 and390x844; narrow viewport is not mobile-device emulation. HTTP(S) blocked; none was requested.", "Ordinary Add/View pointer actions and Space-key activation were tested. No candidate source/state edits or synthesized missing logo.", "The local logo image was verified absent and failed to load with ERR_FILE_NOT_FOUND. No image-generation attempt is present among the recorded tools.", "Targeted behavior and visual review, not exhaustive accessibility or usability testing. Official gate and model classification remain unchanged.", "Performance under the observed Oct22 power/memory-clock restriction is separate from quality." ], "runDir": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/postmortem.md", "nativeTokensPerSec": 4.06, "nativeGeneratedTokens": 4620, "promptTokensPerSec": 202.97, "usageSource": "engine-log", "durationMs": 2439836, "toolCalls": 6, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.594940185546875 }, "peakRssMb": 11960, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-tankcombat", "model": "nanbeige4.2-3b-q4", "scenario": "tankcombat", "repetition": 2, "trialId": "tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu", "journal": "N5", "journalRoot": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fourth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T12:34:11.754Z", "finishedAt": "2026-10-22T14:15:10.209Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,800 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 3.5, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "default-html-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "Infrastructure exclusion", "justification": "No saved/captured index.html exists. A conversion returned a transient HTML receipt, but both saves failed (invalid URI, then unwritable root). The run ended at32800/32768 tokens; compaction removed0. No advisory axes or mean are assigned to unavailable output.", "reviewInputBytes": 10761, "protocolDeviation": false, "reviewLimitations": [ "No saved/captured game exists to render or judge. The transient conversion receipt does not provide the converted HTML bytes.", "No conversion or save was retried, no candidate was created or repaired, and original infrastructure classification/fixed score remain unchanged.", "The combined selected diagnostic text stays below20KiB. Model identity was known.", "Model identity was known during review." ], "runDir": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/783572a4d1.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/postmortem.md", "nativeTokensPerSec": 6.71, "nativeGeneratedTokens": 37357, "promptTokensPerSec": 634.21, "usageSource": "engine-log", "durationMs": 6058455, "toolCalls": 9, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.73858642578125 }, "peakRssMb": 10194, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-schema-migration", "model": "nanbeige4.2-3b-q4", "scenario": "schema-migration", "repetition": 2, "trialId": "schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6", "journal": "N6", "journalRoot": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fifth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T14:50:20.950Z", "finishedAt": "2026-10-22T14:54:51.582Z", "success": false, "failureClass": "model", "reason": "repair-aborted: riley/e75720ba 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.", "modelEligible": true, "fixedComposite": 3.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 2, "robustness": 3, "scopeDiscipline": 4 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Name splitting preserves records in six probes, but duplicate handler exports prevent import and migration return typing fails. CreateUserInput wrongly requires id. Tests and MIGRATION.md are missing. Package/config stay unchanged; the migration remains incomplete.", "reviewInputBytes": 16031, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Only reviewed production exports and advisory in-memory typechecks/probes were executed; no candidate writes or official gate rerun.", "Missing requested files are recorded explicitly; seeded fixtures are grounding, not model output.", "Combined selected text includes grounding and compact diagnostics and is below20KiB. Model identity was known." ], "runDir": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/dd0393ffb3.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/postmortem.md", "nativeTokensPerSec": 58.05, "nativeGeneratedTokens": 5674, "promptTokensPerSec": 2023.1, "usageSource": "engine-log", "durationMs": 270632, "toolCalls": 73, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 7, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.7271728515625 }, "peakRssMb": 11094, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-failing-tests-spec", "model": "nanbeige4.2-3b-q4", "scenario": "failing-tests-spec", "repetition": 2, "trialId": "failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx", "journal": "N6", "journalRoot": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fifth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T14:55:44.659Z", "finishedAt": "2026-10-22T14:57:55.432Z", "success": false, "failureClass": "model", "reason": "repair-aborted: priya/be9d37f6 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.", "modelEligible": true, "fixedComposite": 4.5, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 3, "robustness": 1, "scopeDiscipline": 5 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Default construction throws. Draft cancellation is absent and expedite ignores its flag. Inherited object keys are accepted as transitions, corrupting state. Assignment to readonly state fails TypeScript. The provided tests/configuration are unchanged.", "reviewInputBytes": 12278, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Only reviewed production exports and advisory in-memory typechecks/probes were executed; no candidate writes or official gate rerun.", "Missing requested files are recorded explicitly; seeded fixtures are grounding, not model output.", "Combined selected text includes grounding and compact diagnostics and is below20KiB. Model identity was known." ], "runDir": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/2210b3e9b5.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/postmortem.md", "nativeTokensPerSec": 55.36, "nativeGeneratedTokens": 3376, "promptTokensPerSec": 1727.87, "usageSource": "engine-log", "durationMs": 130773, "toolCalls": 37, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.6239013671875 }, "peakRssMb": 10723, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-symptom-debug", "model": "nanbeige4.2-3b-q4", "scenario": "symptom-debug", "repetition": 2, "trialId": "symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil", "journal": "N6", "journalRoot": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fifth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T14:58:48.578Z", "finishedAt": "2026-10-22T15:02:52.809Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line cursor fix removes skipped records and preserves the API. All62 advisory pagination, invalid-input and direct-cursor checks pass; acceptance and package files are unchanged. The adjacent comment still incorrectly says to resume one past the cursor.", "reviewInputBytes": 7792, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Only reviewed production exports and advisory in-memory typechecks/probes were executed; no candidate writes or official gate rerun.", "Missing requested files are recorded explicitly; seeded fixtures are grounding, not model output.", "Combined selected text includes grounding and compact diagnostics and is below20KiB. Model identity was known." ], "runDir": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/b91a796d68.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/postmortem.md", "nativeTokensPerSec": 41.98, "nativeGeneratedTokens": 8805, "promptTokensPerSec": 1479.18, "usageSource": "engine-log", "durationMs": 244231, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5140380859375 }, "peakRssMb": 3929, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-data-wrangle", "model": "nanbeige4.2-3b-q4", "scenario": "data-wrangle", "repetition": 2, "trialId": "data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj", "journal": "N6", "journalRoot": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after fifth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T15:03:46.375Z", "finishedAt": "2026-10-22T15:15:27.190Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,837 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 3.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-data-v1", "advisoryAxes": { "dataAccuracy": 0, "completeness": 0, "traceability": 0, "outputUsability": 2 }, "advisoryMean": 0.5, "advisoryEligible": false, "advisoryExclusion": "Preserved context-overflow infrastructure exclusion; diagnostic artifact assessment only.", "justification": "The JSON contains ten invented customer records: none of the expected56 emails or source IDs is represented. Fields and sorting are valid, but the dataset is unusable. The helper splits CSV on commas and would mishandle quoted names. Sources/checker are unchanged.", "reviewInputBytes": 17188, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Candidate scripts were read but not executed; independent normalization uses frozen source CSVs.", "Missing requested files are recorded explicitly; seeded fixtures are grounding, not model output.", "Combined selected text includes grounding and compact diagnostics and is below20KiB. Model identity was known." ], "runDir": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-data-evidence/88bb410c75.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/postmortem.md", "nativeTokensPerSec": 41.91, "nativeGeneratedTokens": 20522, "promptTokensPerSec": 1013.49, "usageSource": "engine-log", "durationMs": 700815, "toolCalls": 71, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 4, "forceFit": 0, "compactStarts": 2, "compactFailed": 2, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.7412109375 }, "peakRssMb": 9916, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-incident-postmortem", "model": "nanbeige4.2-3b-q4", "scenario": "incident-postmortem", "repetition": 2, "trialId": "incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva", "journal": "N7", "journalRoot": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after sixth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T15:54:50.579Z", "finishedAt": "2026-10-22T16:01:59.677Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,816 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 2, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-doc-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "Infrastructure exclusion", "justification": "No postmortem or alternate authored file is saved. All five evidence files are unchanged. The only write was rejected by the factual-support validator; 62 read calls preceded the preserved 32816/32768-token context stop. Both compactions removed zero. Missing output receives no advisory axes or mean.", "reviewInputBytes": 10763, "protocolDeviation": false, "reviewLimitations": [ "No requested document exists; seeded evidence and the rejected write are not scored as delivered output.", "The validator alleged unsupported claims; this review does not establish whether that rejection was correct.", "Infrastructure classification and fixed score remain unchanged. No candidate edits, retries, execution or score inference.", "Model identity was known. Selected diagnostic input is below the combined20KiB cap.", "Model identity was known during review." ], "runDir": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/advisory-document-evidence/37b7545b95.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/postmortem.md", "nativeTokensPerSec": 41.24, "nativeGeneratedTokens": 9903, "promptTokensPerSec": 1000.87, "usageSource": "engine-log", "durationMs": 429098, "toolCalls": 63, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 2, "forceFit": 0, "compactStarts": 2, "compactFailed": 2, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.73687744140625 }, "peakRssMb": 9652, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly", "model": "nanbeige4.2-3b-q4", "scenario": "ops-runbook-anomaly", "repetition": 2, "trialId": "ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r", "journal": "N8", "journalRoot": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after seventh context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T16:08:03.136Z", "finishedAt": "2026-10-22T16:10:19.564Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Both reports correctly record the service count, checksum, config and stale/71-hour backup, then halt without later-step outputs. Checksum, config and backup are read before the next log write, missing per-step recording order. Clear sources and restart guidance make the final reports usable.", "reviewInputBytes": 8488, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct prose and source/dependency/order review; no rendered-document inspection or official gate rerun.", "Only generated final reports are scored; source files and selected diagnostics provide grounding.", "Operation metadata does not establish earlier write contents; rejected drafts were not scored as final artifacts.", "Full logs and recordings are hashes for provenance, not additional textual judging inputs.", "Runtime/context and hardware performance boundaries remain separate; no settings changed." ], "runDir": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/4e011c084f.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/postmortem.md", "nativeTokensPerSec": 41.1, "nativeGeneratedTokens": 3612, "promptTokensPerSec": 1214.47, "usageSource": "engine-log", "durationMs": 136428, "toolCalls": 13, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.596923828125 }, "peakRssMb": 3923, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-plan-and-estimate", "model": "nanbeige4.2-3b-q4", "scenario": "plan-and-estimate", "repetition": 2, "trialId": "plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch", "journal": "N8", "journalRoot": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after seventh context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T16:11:13.835Z", "finishedAt": "2026-10-22T16:12:41.194Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct owners, sources, signed layout before mover booking and explicit move execution. T6 depends only on booking, so it can run before T5 packing or T4 network readiness despite the day-one rule. Workshop coordination ends at a schedule, with no owned build-completion task. The plan is clear but not fully executable.", "reviewInputBytes": 7757, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct prose and source/dependency/order review; no rendered-document inspection or official gate rerun.", "Only generated final reports are scored; source files and selected diagnostics provide grounding.", "Operation metadata does not establish earlier write contents; rejected drafts were not scored as final artifacts.", "Full logs and recordings are hashes for provenance, not additional textual judging inputs.", "Runtime/context and hardware performance boundaries remain separate; no settings changed." ], "runDir": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/550e022825.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/postmortem.md", "nativeTokensPerSec": 45.41, "nativeGeneratedTokens": 2458, "promptTokensPerSec": 1519.03, "usageSource": "engine-log", "durationMs": 87359, "toolCalls": 3, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5179443359375 }, "peakRssMb": 3930, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r2-nanbeige4.2-3b-q4-core-conflict-synthesis", "model": "nanbeige4.2-3b-q4", "scenario": "conflict-synthesis", "repetition": 2, "trialId": "conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll", "journal": "N8", "journalRoot": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after seventh context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T16:13:34.843Z", "finishedAt": "2026-10-22T16:34:43.725Z", "success": false, "failureClass": "infra", "reason": "context overflow: 32814 tokens needed but only 32768 available", "modelEligible": false, "fixedComposite": 3.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 3, "clarity": 4 }, "advisoryMean": 3.5, "advisoryEligible": false, "advisoryExclusion": "Original context-overflow infrastructure classification preserved; captured artifact assessed diagnostically only.", "justification": "Correct winning date,210000EUR budget and Marcus ownership, with cited conflict pairs. The table leaks the old date, questions leak Priya, and the footer repeats all superseded values while claiming quarantine. Q3-to-Q4 and propulsion/thermal questions are ungrounded distractions. Artifact review is diagnostic only.", "reviewInputBytes": 13157, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded direct prose and source/dependency/order review; no rendered-document inspection or official gate rerun.", "Only generated final reports are scored; source files and selected diagnostics provide grounding.", "Operation metadata does not establish earlier write contents; rejected drafts were not scored as final artifacts.", "Full logs and recordings are hashes for provenance, not additional textual judging inputs.", "Runtime/context and hardware performance boundaries remain separate; no settings changed.", "Factual-support validator rejections were observed but not independently reproduced; no correctness conclusion about that validator." ], "runDir": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/9ce34d9ec5.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/postmortem.md", "nativeTokensPerSec": 43.46, "nativeGeneratedTokens": 33258, "promptTokensPerSec": 1297.88, "usageSource": "engine-log", "durationMs": 1268882, "toolCalls": 85, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 6, "forceFit": 0, "compactStarts": 4, "compactFailed": 4, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.786346435546875 }, "peakRssMb": 11947, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-tictactoe", "model": "nanbeige4.2-3b-q4", "scenario": "tictactoe", "repetition": 3, "trialId": "tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n", "journal": "N9", "journalRoot": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after eighth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T17:08:05.937Z", "finishedAt": "2026-10-22T17:28:05.677Z", "success": false, "failureClass": "model", "reason": "repair-aborted: aitor/6698e575 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.", "modelEligible": true, "fixedComposite": 4.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 2, "functionalCompleteness": 0, "codeQuality": 1, "polish": 1 }, "advisoryMean": 1, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The dark shell renders cleanly, but the board has zero cells and no move can be played. Reset leaves it empty. Source queries cells that are never created; unreachable handlers also misuse NodeList/map and win-state data. This is a styled shell without the required game.", "reviewInputBytes": 9139, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity was known; full source and both rendered views inspected.", "Desktop Chromium1200x900 and390x844, not mobile device emulation. External HTTP(S) blocked; no such requests occurred.", "Only ordinary visible controls and prompt replies used; no DOM/state injection, candidate edits, official gate rerun or API judge.", "Targeted interaction checks are not exhaustive; latent source defects are distinguished from observed runtime failures.", "Unused SVG was read as supporting output, not substituted for the missing PNG or rendered as the page logo.", "Full logs are provenance hashes, not extra textual judging inputs; performance and runtime cohorts remain separate." ], "runDir": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/postmortem.md", "nativeTokensPerSec": 46.23, "nativeGeneratedTokens": 30246, "promptTokensPerSec": 1375.83, "usageSource": "engine-log", "durationMs": 1199740, "toolCalls": 41, "redFlags": [], "autoAnswers": 2, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 3, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.677703857421875 }, "peakRssMb": 12415, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-petshop", "model": "nanbeige4.2-3b-q4", "scenario": "petshop", "repetition": 3, "trialId": "petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n", "journal": "N9", "journalRoot": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after eighth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T17:30:26.728Z", "finishedAt": "2026-10-22T17:49:14.365Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 33,246 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 2.9, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 3, "codeQuality": 3, "polish": 2 }, "advisoryMean": 2.75, "advisoryEligible": false, "advisoryExclusion": "Original context-overflow infrastructure classification preserved; captured page assessed diagnostically only.", "justification": "Responsive cards and Add Pet work, but the required PNG logo is missing and the hero button is white on white. Reset throws on assignment to a const array and retains added pets. Contact is a placeholder and cards lack keyboard controls. The unused SVG is not the requested delivered image. Diagnostic-only review.", "reviewInputBytes": 19118, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Model identity was known; full source and both rendered views inspected.", "Desktop Chromium1200x900 and390x844, not mobile device emulation. External HTTP(S) blocked; no such requests occurred.", "Only ordinary visible controls and prompt replies used; no DOM/state injection, candidate edits, official gate rerun or API judge.", "Targeted interaction checks are not exhaustive; latent source defects are distinguished from observed runtime failures.", "Unused SVG was read as supporting output, not substituted for the missing PNG or rendered as the page logo.", "Full logs are provenance hashes, not extra textual judging inputs; performance and runtime cohorts remain separate." ], "runDir": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/postmortem.md", "nativeTokensPerSec": 46.53, "nativeGeneratedTokens": 21200, "promptTokensPerSec": 1411.18, "usageSource": "engine-log", "durationMs": 1127637, "toolCalls": 65, "redFlags": [], "autoAnswers": 1, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 3, "forceFit": 0, "compactStarts": 4, "compactFailed": 4, "firstTurnPrefixOver": 2, "loopHalts": 0, "maxContextFill": 0.908843994140625 }, "peakRssMb": 12838, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-tankcombat", "model": "nanbeige4.2-3b-q4", "scenario": "tankcombat", "repetition": 3, "trialId": "tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb", "journal": "N10", "journalRoot": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after ninth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T18:10:08.232Z", "finishedAt": "2026-10-22T18:21:10.123Z", "success": true, "failureClass": "pass", "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))", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "default-html-v1", "advisoryAxes": { "visualQuality": 3, "functionalCompleteness": 3, "codeQuality": 2, "polish": 2 }, "advisoryMean": 2.5, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Clear start screen, keyboard movement, shooting and responsive canvas, but tanks are plain circles. Bullets travel fixed diagonals rather than toward their aim. Three enemies render while the HUD says two; reset leaves stored score intact. Source uses frame-dependent motion and very long-lived particles. Basic arcade shell with combat defects.", "reviewInputBytes": 17264, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full HTML plus brief/diagnostics/replay inspected; three screenshots actually viewed.", "Ordinary visible Start/keyboard actions only; no hidden state or forced hit/death/restart. Source-only defects distinguished from observed effects.", "Desktop Chromium1200x900 and390x844, not a touch-device test. HTTP(S) blocked; no network requests or page errors.", "No gate rerun, candidate edit or external judge API." ], "runDir": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/postmortem.md", "nativeTokensPerSec": 43.42, "nativeGeneratedTokens": 25753, "promptTokensPerSec": 1382.78, "usageSource": "engine-log", "durationMs": 661891, "toolCalls": 5, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.558746337890625 }, "peakRssMb": 10490, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-schema-migration", "model": "nanbeige4.2-3b-q4", "scenario": "schema-migration", "repetition": 3, "trialId": "schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf", "journal": "N10", "journalRoot": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after ninth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T18:22:33.586Z", "finishedAt": "2026-10-22T18:32:40.956Z", "success": false, "failureClass": "model", "reason": "repair-aborted: riley/c2f19ab0 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.", "modelEligible": true, "fixedComposite": 3.1, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 2, "robustness": 1, "scopeDiscipline": 3 }, "advisoryMean": 2, "advisoryEligible": true, "advisoryExclusion": null, "justification": "All target files exist, but migration truncates multi-word surnames and drops non-name fields. Tests import a nonexistent legacyUserRecord. Store.add now requires an id and overwrites id-less callers under undefined. Source-only compilation passes while its public contract regresses. Documentation incorrectly promises first-space splitting and full preservation.", "reviewInputBytes": 19050, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected sources and unchanged specification/seed/configuration read. Only changed/generated outputs scored.", "Production exports probed through ordinary API inputs; authored tests inspected but not executed. Source compilation and virtual consumer checks are advisory, not official gate replacements.", "Targeted probes are not exhaustive. No candidate file changed.", "Model identity known; runtime,context,performance and rubric cohorts remain separate." ], "runDir": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e6f0947979.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/postmortem.md", "nativeTokensPerSec": 48.92, "nativeGeneratedTokens": 15495, "promptTokensPerSec": 1299.77, "usageSource": "engine-log", "durationMs": 607370, "toolCalls": 132, "redFlags": [ { "gezel": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/sessions/riley", "tool": "(no write_file this turn)", "argsSummary": "content_chars=2844 h2_count=6 preview=\"[Earlier in this conversation, summarized to fit the model context: ## Decisions made - User schema migration requires …\"", "atTurn": 0, "pattern": "prose-as-deliverable", "explanation": "Emitted a long-form structured deliverable (H1 + H2 sections, ≥800 chars) as the assistant's chat content WITHOUT a write_file/write_artifact/append_to_file call in the same turn. Caller will see prose in chat but no file on disk; downstream gezels reading the asker's history won't find an addressable artifact. Symptom of the consultation-mode 'reply in chat' guidance overriding the deliverable shape — fix in the role's about.md (the Researcher template's 'deliverable IS a file' rule is the pattern) and use `ensure_gezel` + `message_gezel` with `expectedDeliverable: {kind:'file'}` for file deliverables." } ], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 1, "midTurn": 7, "forceFit": 0, "compactStarts": 2, "compactFailed": 1, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.7161865234375 }, "peakRssMb": 11853, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-failing-tests-spec", "model": "nanbeige4.2-3b-q4", "scenario": "failing-tests-spec", "repetition": 3, "trialId": "failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf", "journal": "N10", "journalRoot": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after ninth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T18:33:35.460Z", "finishedAt": "2026-10-22T18:35:28.970Z", "success": false, "failureClass": "model", "reason": "repair-aborted: priya/88cfe301 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.", "modelEligible": true, "fixedComposite": 5.1, "fixedBand": "needs-tuning", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 2, "codeQuality": 3, "robustness": 1, "scopeDiscipline": 5 }, "advisoryMean": 2.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The implementation stays in the requested file and compiles, but default construction throws. The expedite flag is ignored; repeat submit/pay and cancel after shipping are wrongly allowed. Inherited property names are treated as events and corrupt state. Explicit-option happy paths work, with correct history/error shape for many cases.", "reviewInputBytes": 12890, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected sources and unchanged specification/seed/configuration read. Only changed/generated outputs scored.", "Production exports probed through ordinary API inputs; authored tests inspected but not executed. Source compilation and virtual consumer checks are advisory, not official gate replacements.", "Targeted probes are not exhaustive. No candidate file changed.", "Model identity known; runtime,context,performance and rubric cohorts remain separate." ], "runDir": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/c77380beb8.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/postmortem.md", "nativeTokensPerSec": 55.6, "nativeGeneratedTokens": 2456, "promptTokensPerSec": 1781.28, "usageSource": "engine-log", "durationMs": 113510, "toolCalls": 31, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 3, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.612091064453125 }, "peakRssMb": 10488, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-symptom-debug", "model": "nanbeige4.2-3b-q4", "scenario": "symptom-debug", "repetition": 3, "trialId": "symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v", "journal": "N10", "journalRoot": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after ninth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T18:36:23.203Z", "finishedAt": "2026-10-22T18:37:33.753Z", "success": true, "failureClass": "pass", "reason": "node accept.mjs prints ALL CASES PASS with accept.mjs untouched (signals: accept-untouched, paginate-changed, accept-passes)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-code-v1", "advisoryAxes": { "functionalCompleteness": 5, "codeQuality": 4, "robustness": 5, "scopeDiscipline": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "The one-line cursor fix removes skipped rows while leaving acceptance and package files untouched. All62 targeted pagination, invalid-input and direct-cursor probes pass. Scope is disciplined and the implementation stays simple; the retained comment still incorrectly says it resumes one past the cursor.", "reviewInputBytes": 6953, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Full selected sources and unchanged specification/seed/configuration read. Only changed/generated outputs scored.", "Production exports probed through ordinary API inputs; authored tests inspected but not executed. Source compilation and virtual consumer checks are advisory, not official gate replacements.", "Targeted probes are not exhaustive. No candidate file changed.", "Model identity known; runtime,context,performance and rubric cohorts remain separate." ], "runDir": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e9987c93b8.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/postmortem.md", "nativeTokensPerSec": 44.95, "nativeGeneratedTokens": 1749, "promptTokensPerSec": 1497.66, "usageSource": "engine-log", "durationMs": 70550, "toolCalls": 2, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5140380859375 }, "peakRssMb": 3925, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-data-wrangle", "model": "nanbeige4.2-3b-q4", "scenario": "data-wrangle", "repetition": 3, "trialId": "data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78", "journal": "N10", "journalRoot": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after ninth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T18:38:27.866Z", "finishedAt": "2026-10-22T18:49:38.482Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,799 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 2, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-data-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "Infrastructure exclusion", "justification": "No out/customers.json exists in the captured project. A2356-byte cleanup helper was saved, but three installed-script calls failed and a later output read returned not found. Raw inputs and checker are unchanged. Context overflow at32799/32768 is preserved; no advisory scores for the absent requested JSON.", "reviewInputBytes": 14159, "protocolDeviation": false, "reviewLimitations": [ "No customer JSON to score. Saved helper read for diagnosis but not executed or substituted for output.", "All four input/checker files match frozen seeds.", "Full source logs are provenance hashes, not additional judging input.", "Model identity was known during review." ], "runDir": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-data-evidence/44098f1793.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/postmortem.md", "nativeTokensPerSec": 40.98, "nativeGeneratedTokens": 22137, "promptTokensPerSec": 909.22, "usageSource": "engine-log", "durationMs": 670616, "toolCalls": 47, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 2, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.696624755859375 }, "peakRssMb": 3943, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-incident-postmortem", "model": "nanbeige4.2-3b-q4", "scenario": "incident-postmortem", "repetition": 3, "trialId": "incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog", "journal": "N11", "journalRoot": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after tenth context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T19:10:46.948Z", "finishedAt": "2026-10-22T19:13:05.166Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 33,023 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 2.5, "fixedBand": "framework-gap", "advisoryStatus": "noArtifact", "rubricId": "campaign-doc-v1", "advisoryAxes": null, "advisoryMean": null, "advisoryEligible": false, "advisoryExclusion": "Infrastructure exclusion", "justification": "No postmortem or alternate authored output exists. All five source files remain unchanged. The trace contains26 read_file calls, one read_files and one stat, with no write attempt. Context overflow33023/32768 and compaction removing zero are preserved. No advisory axes or mean for absent output.", "reviewInputBytes": 5631, "protocolDeviation": false, "reviewLimitations": [ "No requested artifact exists; seeded source evidence is not model output.", "No write was attempted in this trial; do not import the rejected-write behavior from earlier repeats.", "Infrastructure classification/fixed score remain unchanged. No retries, candidate changes, execution, external judge or invented scores.", "Known model identity; selected input below combined20KiB.", "Model identity was known during review." ], "runDir": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/advisory-document-evidence/f5bf43f97a.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/postmortem.md", "nativeTokensPerSec": 40.45, "nativeGeneratedTokens": 2391, "promptTokensPerSec": 952.71, "usageSource": "engine-log", "durationMs": 138218, "toolCalls": 28, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 1, "loopHalts": 0, "maxContextFill": 0.73114013671875 }, "peakRssMb": 3950, "peakGpuMemMb": 8210, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly", "model": "nanbeige4.2-3b-q4", "scenario": "ops-runbook-anomaly", "repetition": 3, "trialId": "ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid", "journal": "N12", "journalRoot": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after eleventh context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T20:05:57.936Z", "finishedAt": "2026-10-22T20:09:49.381Z", "success": true, "failureClass": "pass", "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)", "modelEligible": true, "fixedComposite": 9.5, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 4, "actionability": 5, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Correct ordered final STEP1–4 values and clear expected-versus-observed halt report; downstream outputs are absent. Actual reads reach config/backup before the checksum and before any log write, violating stepwise execution and recording. Repeated source reads and two rejected halt drafts precede the useful final reports.", "reviewInputBytes": 9854, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded prose, source, dependency and operation-order review; no rendered document inspection or official gate rerun.", "Only final generated artifacts are scored. Operation metadata does not establish earlier draft contents.", "Full logs/recording are hashed for provenance, not additional textual judging inputs.", "Factual-support rejections, if observed, were not independently reproduced; no validator correctness conclusion.", "Runtime/context and hardware-performance boundaries remain separate; no settings changed." ], "runDir": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/574fd8b47e.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/postmortem.md", "nativeTokensPerSec": 41.69, "nativeGeneratedTokens": 5498, "promptTokensPerSec": 1276.71, "usageSource": "engine-log", "durationMs": 231445, "toolCalls": 19, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 1, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.547607421875 }, "peakRssMb": 9335, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-plan-and-estimate", "model": "nanbeige4.2-3b-q4", "scenario": "plan-and-estimate", "repetition": 3, "trialId": "plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84", "journal": "N12", "journalRoot": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after eleventh context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T20:10:49.586Z", "finishedAt": "2026-10-22T20:12:28.971Z", "success": true, "failureClass": "pass", "reason": "Plan satisfies structure + workstreams + ceiling (signals: ordered-sections, plan-structure, ws-floor-plan, ws-movers, ws-network, word-ceiling)", "modelEligible": true, "fixedComposite": 10, "fixedBand": "ship-ready", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 5, "functionalCompleteness": 5, "actionability": 4, "clarity": 5 }, "advisoryMean": 4.75, "advisoryEligible": true, "advisoryExclusion": null, "justification": "Complete eight-task plan with internal owners, signed layout before main mover booking, actual workshop completion and relocation gated on network readiness. Network/print work goes to workshop/design owners instead of IT. Backup mover prebooking in Risks lacks an explicit sign-off dependency. Clear and useful, with those execution ambiguities.", "reviewInputBytes": 8183, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded prose, source, dependency and operation-order review; no rendered document inspection or official gate rerun.", "Only final generated artifacts are scored. Operation metadata does not establish earlier draft contents.", "Full logs/recording are hashed for provenance, not additional textual judging inputs.", "Factual-support rejections, if observed, were not independently reproduced; no validator correctness conclusion.", "Runtime/context and hardware-performance boundaries remain separate; no settings changed." ], "runDir": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/558843afd6.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/postmortem.md", "nativeTokensPerSec": 43.41, "nativeGeneratedTokens": 2849, "promptTokensPerSec": 1543.61, "usageSource": "engine-log", "durationMs": 99385, "toolCalls": 1, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 0, "forceFit": 0, "compactStarts": 0, "compactFailed": 0, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.5179443359375 }, "peakRssMb": 3952, "peakGpuMemMb": 8208, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true }, { "cellId": "r3-nanbeige4.2-3b-q4-core-conflict-synthesis", "model": "nanbeige4.2-3b-q4", "scenario": "conflict-synthesis", "repetition": 3, "trialId": "conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn", "journal": "N12", "journalRoot": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22", "runtimeCohort": "Corrected runtime / 32K Nanbeige after eleventh context stop (Oct22)", "contextTokens": 32768, "slots": 1, "engineContext": { "launches": 1, "grantedPerSlotTokens": 32768, "totalTokens": 32768, "slots": 1, "kvCacheType": "f16", "launchModel": "nanbeige4.2-3b-q4" }, "startedAt": "2026-10-22T20:13:26.475Z", "finishedAt": "2026-10-22T20:21:06.787Z", "success": false, "failureClass": "infra", "reason": "On-device model ran out of working memory: 32,794 tokens needed but only 32,768 available. Try asking a narrower question, or ask me to search for a specific piece of info rather than fetching a whole page.", "modelEligible": false, "fixedComposite": 3.6, "fixedBand": "framework-gap", "advisoryStatus": "judged", "rubricId": "campaign-doc-v1", "advisoryAxes": { "factualAccuracy": 4, "functionalCompleteness": 3, "actionability": 3, "clarity": 2 }, "advisoryMean": 3, "advisoryEligible": false, "advisoryExclusion": "Original context-overflow infrastructure classification preserved; captured artifact reviewed diagnostically only.", "justification": "Correct winning date, budget, DRI and allocations, but the file contains two complete briefs. Both question lists repeat the superseded budget; the second plan leaks Priya. It treats absent cadence as superseded, invents finance headcount and labels a historical budget difference a shortfall. Duplicate sections and false quarantine claims weaken it. Diagnostic only.", "reviewInputBytes": 14841, "protocolDeviation": false, "reviewLimitations": [ "Model identity was known before artifact review; this assessment is not fully blind.", "Unblinded prose, source, dependency and operation-order review; no rendered document inspection or official gate rerun.", "Only final generated artifacts are scored. Operation metadata does not establish earlier draft contents.", "Full logs/recording are hashed for provenance, not additional textual judging inputs.", "Factual-support rejections, if observed, were not independently reproduced; no validator correctness conclusion.", "Runtime/context and hardware-performance boundaries remain separate; no settings changed." ], "runDir": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn", "judgmentPath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/codex-inline-judge.json", "evidencePath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/b317f0577b.json", "postmortemPath": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/postmortem.md", "nativeTokensPerSec": 42.39, "nativeGeneratedTokens": 12815, "promptTokensPerSec": 1201.48, "usageSource": "engine-log", "durationMs": 460312, "toolCalls": 46, "redFlags": [], "autoAnswers": 0, "missingRoles": [], "compaction": { "observable": true, "betweenTurn": 0, "midTurn": 2, "forceFit": 0, "compactStarts": 1, "compactFailed": 1, "firstTurnPrefixOver": 0, "loopHalts": 0, "maxContextFill": 0.75640869140625 }, "peakRssMb": 11962, "peakGpuMemMb": 8212, "runId": "2026-10-20-rtx5070ti-nanbeige-core-32k", "includedInScorecard": true } ], "evidenceHashes": [ { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/result.json", "bytes": 848, "sha256": "279f3f5d2a784135d224316c59ee3a2dc9cbd71ad20ca1db7fd0d3e9fa32e7c4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/score.json", "bytes": 3088, "sha256": "bb23c98980d4302ae49effbe6b98b90668f2440e883d1f4ab8b61202b8de3814", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/metrics.json", "bytes": 2370, "sha256": "67920b67b23965a65c78799a06b4bba38ddb7c8c3320efad0b4da1e18e2e00a0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/facts.json", "bytes": 11691, "sha256": "e7e7cadd223df0209af94e7b9a7d132b829f4e4e019782bb33c1293176eb2567", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/codex-inline-judge.json", "bytes": 1651, "sha256": "53bb1c1a2ec38610062b519f1fdb13a6f0dcfbb24c5e07e28a92c7891404e8d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T02-24-29-313Z-cw1u/postmortem.md", "bytes": 7635, "sha256": "c0b0b3cc2f2df6a16f12eec9365e492e1db1119c32e903aed756219589a38486", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/5ccab49fa8.json", "bytes": 1452, "sha256": "3ceab09dff3ee7a952d84c807653cfe8cc373625b37036172efd918cf445e7f6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/5ccab49fa8-desktop.png", "bytes": 10892, "sha256": "8986923ac1ab08c18f9d53cd8dfe2de2217159140e81f958e689f79637805bd6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/5ccab49fa8-mobile.png", "bytes": 9748, "sha256": "f3511924f4823c8e23973425b6cc7161b0d954d923a55bf4e6fa7e2271bf3834", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/result.json", "bytes": 847, "sha256": "5cea30d193b769db6e0f68a142210aaa1a7e366baeae5364f66013de51f7f2f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/score.json", "bytes": 3088, "sha256": "59c664479e619b69a36ef2677f2bb01419adddf7b00aedc717735ac217d893b2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/metrics.json", "bytes": 2332, "sha256": "5d2559616f698b5063edca60dccaba9d2983dae9d364f46b19133fa8b62486c0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/facts.json", "bytes": 12224, "sha256": "56e6b927c0a5ae679b71b856c5619e010869fe4ca53c58dc3baf8e5f399b33e8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/codex-inline-judge.json", "bytes": 1673, "sha256": "e85d2e348eb29e401d68279bae571fc2dc333d083e54f03654ed55dd6a044fe0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T02-27-32-991Z-rqab/postmortem.md", "bytes": 7657, "sha256": "5d5e65e06441ce5a779982b7653adda700abb5f7a04ecc93c8bfb29daf7de26b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/7847a7a070.json", "bytes": 1452, "sha256": "92afbfc278357b73a6617db5794e3fa5b9528cccc98ed91dbb165e55fcd2b8eb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/7847a7a070-desktop.png", "bytes": 8133, "sha256": "3b63b553a2743430863d870a50d219546fef44dacd246641ef799c7f3bc04dfe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/7847a7a070-mobile.png", "bytes": 7070, "sha256": "226a2da675b27661e988567b2b1e363e56da4b4c889cba6c1a5f9faa77ca200f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/result.json", "bytes": 1247, "sha256": "4f6124e5abad7ce97ffa2ccf2c6d7749aaae619646edd4a48404d22ca6adc8cb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/score.json", "bytes": 4724, "sha256": "eec7e8e41699c2156d7efeba6744cbf20d579770f97f7da1f4ca2be8483c869b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/metrics.json", "bytes": 2372, "sha256": "cf23792db270069d49178b2d6c536d300b624dd3a5e28541fa2f4faa1a255eaa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/facts.json", "bytes": 14055, "sha256": "7e198f95ef210a8a95d93990798cb973e815b062502e119b7012f5aad546c06f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/codex-inline-judge.json", "bytes": 1685, "sha256": "a812ba644128c394ec7f22a98e89062bbc83498a6266ea91aaf8926b9c5e97cb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T02-30-22-563Z-wfod/postmortem.md", "bytes": 9942, "sha256": "b00b0f675d50c0baa5d27477e472baa880eb36a7353ecf20b10b8115bb006d65", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/970ef8687a.json", "bytes": 2238, "sha256": "9935e89a132144df37853c71a792244c44e4bea3c7950a0b8c22e302e8a59cd2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/970ef8687a-desktop.png", "bytes": 9114, "sha256": "28d13736aa0703778dc1953bd1a5377fa5bdeb3bb659251e48b9bc77dbf09f3e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/970ef8687a-mobile.png", "bytes": 7999, "sha256": "c740bbdd7ea19bdb194bd438f9f5f0d376277caebe0554db8976b5474d98088c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/result.json", "bytes": 853, "sha256": "2ce7f7a2efff7147101af2834a92330d25f62f424ef589bf91a569c3507f4cbe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/score.json", "bytes": 3098, "sha256": "6cf0d73ac91b161aecc27f97e749d5071ca2f9b4ee4b2a272462b28e17523b41", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/metrics.json", "bytes": 2373, "sha256": "15e44b07db13ac4431f960899be838479c6cf2dc86c5503b43acfab887367a06", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/facts.json", "bytes": 13157, "sha256": "536d9cec5ee725b267b3de2e583c0cea21bd633cf7e93e0f8bd7e0a765f8d217", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/codex-inline-judge.json", "bytes": 1869, "sha256": "91f0f72b773e892b83e8506898997cb20ad44422680c87341e8b902fde64220a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T03-31-25-065Z-1y85/postmortem.md", "bytes": 7704, "sha256": "2eed66b32b0568a3011781e17ef270c091dd26d8cce3115b5b3bee62eed88f9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/78cefb44ef.json", "bytes": 1958, "sha256": "2d98fb0caf1bdb756e47ef153ecf55ebfeeac938d3897ca0cfd8fd521cbf1c20", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/78cefb44ef-desktop.png", "bytes": 10916, "sha256": "d4e9a66a25ee96a399dd3b54d0b5f180b5aed27338ac65a8b9b3bfaefa43a32c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/78cefb44ef-mobile.png", "bytes": 9833, "sha256": "5a55e36f8228909ebb27498774194e153f5d998c083239074fb3b00676798549", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/result.json", "bytes": 847, "sha256": "5743616ee6f08ab82987e0a156297c66c34eabb740531a4fff416c462eb680af", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/score.json", "bytes": 3097, "sha256": "c522c3b22da1b436c1362173fd05387e3c091d72d4a62708eab30c8da5d92ad7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/metrics.json", "bytes": 2372, "sha256": "e3f9478013e20e66d24ebaf78044eeefdf4057ab5c4bbcddce33f4a8d6685fbc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/facts.json", "bytes": 13755, "sha256": "a54377f70faa9e96ad9a2a4d10498786e2a771fdc8c0de8dfe858bd22e80bb7c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/codex-inline-judge.json", "bytes": 1854, "sha256": "922a27cd42916573226a4d1ee645028af64a12b02d38c1811cf257fff9e7c2d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T03-36-21-214Z-pf0c/postmortem.md", "bytes": 7684, "sha256": "5c8cb0d2ea41a646df1e718960f004cd1fd1576270d430aea275287a97ce8e16", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/fe65af7289.json", "bytes": 2103, "sha256": "2610736b8915d497552c8f91474f36b95083d5218bde8e47bc595078db8e67ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/fe65af7289-desktop.png", "bytes": 12685, "sha256": "9d58cb7f2fb22eb1b4735c3fee8b49150e62f2982d103f3c4df5bec093e7f624", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/fe65af7289-mobile.png", "bytes": 11387, "sha256": "90a690021c17bcc71065b9cc609501077c5d1d83f92f267f53dde1c856595706", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/result.json", "bytes": 919, "sha256": "799eb5c0797b356ef190a345769b6586c28ea7ef4cd47b92a8b541c1e65b3122", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/score.json", "bytes": 3119, "sha256": "05f1128fd90ed3faf171c47dca8cf5145b03da49d9c6401128e9b0ee14284a02", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/metrics.json", "bytes": 2402, "sha256": "333343a30db7f805e5cd65331a75682bce12739f3577e4e9a9fe5ce63e0e0478", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/facts.json", "bytes": 12589, "sha256": "ab3682ecb4759fde952270fae60905e83145597fbdf71242efc44739306d8132", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/codex-inline-judge.json", "bytes": 1908, "sha256": "71a9ca877392492b9fdcfbf3f5244e213d88182a248319846dd9cdd416facb9c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T03-40-52-396Z-yg0u/postmortem.md", "bytes": 7687, "sha256": "8ab638f4f66352ab675103a177a1681fc7e6be06233b115a62383666933a42a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/a3baf5ab15.json", "bytes": 2221, "sha256": "a3f5b52a791ae7b4d2c607d7650677fe0c024b93dacec9b21527760edeeb1b33", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/a3baf5ab15-desktop.png", "bytes": 12095, "sha256": "4ddf9cdf9087eb63e060efb7310502a30c94b28a1ce75546d78c19e3da710597", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/a3baf5ab15-mobile.png", "bytes": 10301, "sha256": "52730339984bf832ac66b5c114ab902e1cbf77da8d6b5f32583de32c3c0e9fe3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/result.json", "bytes": 854, "sha256": "bc291fa98f0774e63d53ee3e5bf145474e1558d379f61f31e2ad28f975ad801c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/score.json", "bytes": 3093, "sha256": "b5170b70597ead4653572138fcfe4a541e6ed2bf857ad1609e54d836e85b62cd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/metrics.json", "bytes": 2372, "sha256": "c76b6514b0cb23929f2e53dfa75c31ae44bebd2a3d6fae3f4febedfc92e77ef6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/facts.json", "bytes": 12324, "sha256": "02733cd7da8a8c1cc622571a352421aec465a6196999d04e9f51e25f3f89bd47", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/codex-inline-judge.json", "bytes": 1877, "sha256": "dd7e3186e6fb7836aabfeab44655be47704aed9e6e539f533707e9456ecb716f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T03-48-49-115Z-qov1/postmortem.md", "bytes": 7695, "sha256": "590c0844e6a37f8b5f35462df16874f659be32da40ddaa8f25961a5166faa8c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/29df2a5d56.json", "bytes": 2133, "sha256": "01ff8320d32ad30eb07e05849c873295dde517fdfd649afd7f6552494c714919", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/29df2a5d56-desktop.png", "bytes": 11845, "sha256": "a4bd205e11fb8d920e93a322893096bb5d9af45b7ec42a259d5b85c338ce8068", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/29df2a5d56-mobile.png", "bytes": 10204, "sha256": "c897394b0322b3efef88d7e7f19050ea193c1b8872f3dc659e82c0ca9c164337", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/result.json", "bytes": 859, "sha256": "8c4162b2ac43c0c343e7bfb66e3062eb394547eff2bc86d41d6b035e0a95ec31", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/score.json", "bytes": 3609, "sha256": "2f10e85543d45390e091f3ce1113968fc54d08b89f9eb25c0a01266d2ae38cc2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/metrics.json", "bytes": 2376, "sha256": "838e2afe0f76e4df68a4098e19bfdf7450e3aa8f599a452bd71697d1b95b3f39", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/facts.json", "bytes": 14187, "sha256": "b42b7094fad56ce55fbe466d9329a4fe1df3f08351433d6905659156502fb063", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/codex-inline-judge.json", "bytes": 2046, "sha256": "42d8c25081c1722891b61c2af65141af03645d715b27c1a55c0f7c8dd50f177e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-20T04-00-51-725Z-7y5e/postmortem.md", "bytes": 8247, "sha256": "f104edae503460bff615b9d329aa0942f4f89f21d5b221519f2d5cd7909cabd0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/9f799cc233.json", "bytes": 1445, "sha256": "18155251bd8d31d9758d832a55760d917ad553a9959939e63a4799a6aff0b614", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/9f799cc233-desktop.png", "bytes": 57927, "sha256": "c7047880d08a35990354096502907a7a213a8ac57ff37b0c1fdff3537efb505d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/9f799cc233-mobile.png", "bytes": 53935, "sha256": "a820e62d7e6419111870ee8d9810a2c144aceb2b71a0fe779023bf31f83463ae", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/result.json", "bytes": 853, "sha256": "5fc3d2e0b14863d9f6670d0bf687842685b04051645784cf97032c205faa1126", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/score.json", "bytes": 3668, "sha256": "d2f78eeb87f7607e31e3b40ab1385af8293bf0ec21770acda012068713b22f22", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/metrics.json", "bytes": 2381, "sha256": "338db7902bac217b7d9de47742f94a2dcb7598535ee963c60cbbe39ce54ae168", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/facts.json", "bytes": 14031, "sha256": "51841811a349c7272c58cdf0b7758ece89c268509b8eb20b605f051e96368bd5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/codex-inline-judge.json", "bytes": 2054, "sha256": "9a9a801b3fb690ddaaf3582af9b4d350dc3a78ece5b01b1c8f5563f0aa2fceb8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T04-04-29-763Z-r0qg/postmortem.md", "bytes": 8129, "sha256": "118fd3f2d5aee22377d58151b7dd893cfb68ac1385fcee162de11db50a15c8f2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/befd0b9f84.json", "bytes": 2978, "sha256": "4141c6bef1a496a79342df8bc47ec5b25779f933271f586af7032a0274750477", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/befd0b9f84-desktop.png", "bytes": 104807, "sha256": "eaa3c452ccb723cdbd83d543ebf4697af0b6ce773e9bf1b0ccf66a573a200d65", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/befd0b9f84-mobile.png", "bytes": 87169, "sha256": "eb2a224762cce8196d394577d14657646bd30491b88ef8bc1abc25ba6e0bba96", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/result.json", "bytes": 886, "sha256": "4e2dac6090508e906d9e6d6d5a218a1aaa171544a2cce810a11abc624c1816ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/score.json", "bytes": 3666, "sha256": "36fb7f3d96336c5a584199c2e79c50d26e6596ab2fe632817dde9d6df7b026a1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/metrics.json", "bytes": 2336, "sha256": "6dc31b99c6e3ab66d5d1f563871afe85393f58ccf3c5fa1b253e9a28600325fb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/facts.json", "bytes": 13537, "sha256": "c1dfab0935f13c280769140d05f2b1bb1866cda69260b62e8f84b006073f8223", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/codex-inline-judge.json", "bytes": 2085, "sha256": "66cc1f2ea529aa220de6ba507723c479f5ba45a40a16355623c3b33f8c6fdb4c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T04-48-11-959Z-n4ec/postmortem.md", "bytes": 8184, "sha256": "b426b8d6cdc7cbc0a27f63f2ac4f2a404f636bf86ffc12c616eda09bb2baea05", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/33e82f57c1.json", "bytes": 3147, "sha256": "7af0a3b321bd850eea7702c75d98e52965452d018d60bde8ff826bb6557b7418", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/33e82f57c1-desktop.png", "bytes": 147554, "sha256": "ec76f12e4cb3d4f54490b1f543c485c33fc3e9d474fc7f538c9af7d1d061ee0c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/33e82f57c1-mobile.png", "bytes": 114364, "sha256": "ce510440de520915315370b11df119e9fab2f6903b72194d5418cecc7afe104c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/result.json", "bytes": 1075, "sha256": "3f64dc5e1eea0abd58b4fa65f6e2fd19813d3166b2736e9fc145bf9a5beaf098", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/score.json", "bytes": 4419, "sha256": "15e32c446b6c73362309f9b626e0c971047c607b9ed64753278d46f6dc14c7b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/metrics.json", "bytes": 2387, "sha256": "485c7c1801fabc7270bc77a021dccad55388fbdc2706077a7c01c7cd3a63a3df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/facts.json", "bytes": 15768, "sha256": "f2baafcba8c214e1c4e36bdad2e3fd4b7d4a58880a67fc79652932261307b4ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/codex-inline-judge.json", "bytes": 2643, "sha256": "14ea6c732156d0e7fa63cc194cbb38e5e4543f5dac2aea17d2f30adc6f10b9d1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T04-53-26-550Z-del9/postmortem.md", "bytes": 9759, "sha256": "29fdb3f0d3d6ef9e12e8dac5379f4ab2e3cd40a76412d7cd0db3e3b1f022d095", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/cb442c4b5a.json", "bytes": 4226, "sha256": "ad763454ca5e88733687a2a5345d6b813cb04331ab7439093f94bdd0b9228da6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/cb442c4b5a-desktop.png", "bytes": 97734, "sha256": "7eb70ce4fe0c851472ff5eae7d254de5323b753781475bb14ff4dfe3b032896d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/cb442c4b5a-mobile.png", "bytes": 67556, "sha256": "3a4d08cb9c541bc4fdb4b93e8a544bae770b7bd2a5c74093cb048b761a37e5d0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/result.json", "bytes": 925, "sha256": "6a8c79d5a4309b28ce5ef22557209d0b9bab5d9365427289445ba47cd55b5293", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/score.json", "bytes": 3614, "sha256": "15d82b64c1fa9db7208f67a1a7cccb2775c72c62148b5d6eda3712f401f20730", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/metrics.json", "bytes": 2373, "sha256": "60b400295158edf887b3fbd07079a66080d810edd3ef592632953cf392b7f19f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/facts.json", "bytes": 14717, "sha256": "f790aa6453dd57a8f978f9c047ec793a325b44281386374b83888bbcaffba807", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/codex-inline-judge.json", "bytes": 3178, "sha256": "77a99c8d150e63f4970b3b3e77ffc3d7724dbab47513af2d5e7002285864b464", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T05-03-48-706Z-fsc1/postmortem.md", "bytes": 8716, "sha256": "9ff776ed886880e0e98bdcf1fb3474c4e66873cb3510cdc02901c864c288bacd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/e957bf320e.json", "bytes": 8389, "sha256": "12747b192fd568209d075199b51e5ed97acd0c9aec482eece2dd68b56a06982b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/e957bf320e-desktop.png", "bytes": 287274, "sha256": "792c1033496ef6a87ae608ec57d6b77a4d15fa4313ffc0f54f88319ce8b1a3b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/e957bf320e-mobile.png", "bytes": 254583, "sha256": "c14c75b0391f4103c343192fdc350f25b1e033681a52a8cabca9a27c478be689", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/result.json", "bytes": 1143, "sha256": "258c848f16538ea6fdb481b540d2e3bf1ad64e0b948d2a804c51c97071529774", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/score.json", "bytes": 4392, "sha256": "a4fbcce2f29874424e056aad85870a2d7adb05a7be13b3ec753677a08f6d6350", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/metrics.json", "bytes": 2383, "sha256": "39aec4c50117116dba73aed50994f046d293b3f344a81f33db6464bf78bbc004", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/facts.json", "bytes": 13573, "sha256": "1f89115c9e37acb0df76a94539ff7cfa24f0d295d2d0258eef4afc7132847ffc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/codex-inline-judge.json", "bytes": 2397, "sha256": "9d8ef66b8b0ea6fd10583dd2652bec57eae59cebc195cbcbfad2f584ea397ebe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T05-40-18-401Z-9n4g/postmortem.md", "bytes": 9675, "sha256": "c24206fbc62aadce6532d516adc274929f174baa70baf5dff9fdd3e028925e8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/ce09aa4981.json", "bytes": 3760, "sha256": "fa5927a0c928cb0754cac8944b7dc91b8d25188d25669e0189949360ad39d7d5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/ce09aa4981-desktop.png", "bytes": 247519, "sha256": "66c87b312143a77d8f081b24108f29bc6331a6feea7c5327165924784c05547a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/ce09aa4981-mobile.png", "bytes": 150858, "sha256": "6d295940aed24019198e3c700fe49b7fcea74a79f4e8148730dc2897dec8ed9b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/result.json", "bytes": 854, "sha256": "3a603bc8c9b6ab626d73cc40d9f505c6d8bc8980a9dc9cb9b6aeb71bee4eb5a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/score.json", "bytes": 3569, "sha256": "fb6dfcd910bba332f3363467f38774899fc06410ecbf6f6b1939205e3f50e888", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/metrics.json", "bytes": 2376, "sha256": "4e0b657e8151f466bd7ad83e4139b15b21abc56a16b3c915842e179875bd371a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/facts.json", "bytes": 13849, "sha256": "83c9dd3943e40c91cda2ba5f89d8f23885173013df6a9bae760a8bae8bf247b8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/codex-inline-judge.json", "bytes": 2387, "sha256": "136fede43f3b6c21ebafa69236180aa85b3abd99e8b2295c5bc2806885a3e5d0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-20T06-32-43-734Z-j9uv/postmortem.md", "bytes": 8108, "sha256": "e1490749899a18c4746ee288c0d8e46c7355ea78ecfd1b90980fc1dd255e44df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/4e8ae2df8a.json", "bytes": 3331, "sha256": "a3f3044b05a1b1c0dcc10228a50c60fba90cd934db5b1b9be64e272ca4e2cac7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/4e8ae2df8a-desktop.png", "bytes": 78375, "sha256": "ab7d937f39b6f0c0b7801c5221bbaaa763d3749806a6eec9741a0ab5c650c399", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/4e8ae2df8a-mobile.png", "bytes": 70637, "sha256": "c029bbb20467d773fdca31edeb5f309aadde750e09562e9f61a0c24e0347a21f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/result.json", "bytes": 911, "sha256": "643a28427db40af2ac0f7e318188758ec4084e1cc4522918f83fb69e7cde317e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/score.json", "bytes": 3200, "sha256": "729617a70320078dcd0dd725aaac0b892aa229e6cc38d2539d58b96e66b22f1e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/metrics.json", "bytes": 2365, "sha256": "e6405646369466357468835be0ebe79060c5b496d5a735eb1d3347178689d2c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/facts.json", "bytes": 12412, "sha256": "1ea8782f6740609d407bcf8094f0351a44f1193695098cbbe2c6cf4d58c32d5d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/codex-inline-judge.json", "bytes": 2437, "sha256": "5a1e9ee4204885181cca031bcd66c03f9c35c7f34f9258cbc1700a08c775cf28", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-20T06-38-01-836Z-ufiq/postmortem.md", "bytes": 7596, "sha256": "aa2caa58d79d6360ca107192cf8889409ef6e5713f6a4e013135c0ead31fca1c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/0466a88572.json", "bytes": 4054, "sha256": "e16aa0cb0e1736a5b9d9e50fb2969b2c53ae3a0b210b62c88499254fb38acfc9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/0466a88572-desktop.png", "bytes": 7436, "sha256": "e2a2a6466bb7372d9e1ec8426f1f7d44c7bad00a87f0eb8b4687378ffc694af1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/0466a88572-mobile.png", "bytes": 5716, "sha256": "7dd4365d2498e89844a9fef66b0499f1b808c99ba7950392998266b2311426ec", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/result.json", "bytes": 921, "sha256": "e32f61873618641deb92a1561bc45fff09bcfee0fd0d31e3537393c62211879e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/score.json", "bytes": 3202, "sha256": "f5d33f4e04ac18bad9025aa14f93d7a7feaa871deb2e88b5ed94e2072aba26c5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/metrics.json", "bytes": 1310, "sha256": "676b6a4f1f5fe796d174a3d444aa7647b75d335e5018d0541005bf27901cab9b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/facts.json", "bytes": 11620, "sha256": "09668eb0c57f4a243a060584da7026b98fc5a2e92f0547e18eb736459ca35928", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/codex-inline-judge.json", "bytes": 2450, "sha256": "bde7618ef1f60b1bf234df70cb76b6afd9f56df1b6f267bdf21f3a13cc445fa5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-20T08-33-10-914Z-g2n0/postmortem.md", "bytes": 7580, "sha256": "55f2f792611eb18e0e132a54228573e9997b09908fb332512fb4efc3b254436a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/c37f5bcf48.json", "bytes": 5383, "sha256": "b90070d1c5a3aa2b9b1c130226909aa0d5950acf47511d374d047f32a63e280d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/c37f5bcf48-desktop.png", "bytes": 30156, "sha256": "76d0ae18fc71fb063a731ada578cbb58c047e44aaf90c8da4fc66f690475bba9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/c37f5bcf48-mobile.png", "bytes": 18984, "sha256": "f0f9e1640e5ea29c5edce15bdc99fcc395e9a9e05e5afb8b60323d198875f6c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/result.json", "bytes": 909, "sha256": "4e107d89ee34fa4b26de5d718d9ab7a8d1ffa54d00ebd5a1cc11a4d594c4afb5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/score.json", "bytes": 3196, "sha256": "3881a1687ad997ba988364e605065a81c5934a2f463558363e02073e0be4225a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/metrics.json", "bytes": 2369, "sha256": "d5b0b24dc2f6b8e85a924a1ab01a89531adb0ee82961d5d0c47db0da21b5ac8e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/facts.json", "bytes": 12251, "sha256": "b31338dc306b7c04bbbf95383a6f3fda76758f956cf0be6ddaa0481754b4237e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/codex-inline-judge.json", "bytes": 2428, "sha256": "e1342867608d5cd40fe3fdf373a70ed02ad81a6c78fa535ca77a8a76aa328456", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-20T08-47-57-751Z-76n2/postmortem.md", "bytes": 7572, "sha256": "06f0d2c0969c5047c8ce5ae5b1b456a147b6b88470cfb873b56c602c238e3db8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/6c189d9c1e.json", "bytes": 2962, "sha256": "c13bc8352b3589bff5829cf14dace34832f80fcab60ad7eb94fa1d5d8b42563e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/6c189d9c1e-desktop.png", "bytes": 10955, "sha256": "5c5b0c1877132c88ab0dc77b196b2fde2075afbf24edd488f25ad7e7fe083207", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/6c189d9c1e-mobile.png", "bytes": 6079, "sha256": "2448cab42d4b87a298d79dc622e316b17446e80283892fdff2dbbd673d3836e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/result.json", "bytes": 982, "sha256": "b3cc8dbb65d542723eb7d12017c7990e96a50fccea22e86ff72923ccbcdc7468", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/score.json", "bytes": 3226, "sha256": "264eaa3e0c65f3abb94a1c5007b474f1752202120c72c8937f66fd6945dda2bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/metrics.json", "bytes": 2398, "sha256": "3598085d90bb46a342372d5ded3868017690511fd2165d96bcb9d3e91f1fd750", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/facts.json", "bytes": 12670, "sha256": "93526fdd0e599251516dd4e6a7734867a6e7bfcd086ea8d3fd0b8ec0098c08a7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/codex-inline-judge.json", "bytes": 2529, "sha256": "dae3c5ad94a08e19081d2b583fe165bb823c511293bfe2050632f6f363bc6e88", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-20T08-50-28-106Z-8hwm/postmortem.md", "bytes": 7613, "sha256": "6988ade82583338148712536280089dbe3f89b631968484514379be5768f6c7b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/80664c8d69.json", "bytes": 3183, "sha256": "2bdd89e389460b574fa3588f32a3561dc14d7853ac7ce574d32f79fef7279e7f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/80664c8d69-desktop.png", "bytes": 21508, "sha256": "775cf4964609e738def0f06d8078da95f4a51ac4792ef070d447e88bcc707891", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/80664c8d69-mobile.png", "bytes": 13295, "sha256": "77cdceaf4256bf53a38565b6a6ecf45734aaea4d8c720a119e9d35460240e0ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/workspace/tank-combat-arcade/index.html", "bytes": 9947, "sha256": "7e2cedbf5a5918d072b048ec1dd09e1a651b06b85952fd5d62bfc8e5fdfd28d6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/result.json", "bytes": 1162, "sha256": "5a9978467004db667f932669c9339042ff179053dc3a5284ba6a4dc3061279fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/score.json", "bytes": 4491, "sha256": "d0105b8b8d346f1fbe086468eb0ed87cd762af5b796ae3c2d41a7cd7a727b385", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/metrics.json", "bytes": 2375, "sha256": "3265e9eb95e8f537cf0269aee2c047a22dfb77d12d2eddc1f0027054c54b6e36", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/facts.json", "bytes": 12920, "sha256": "14181e6c2b18263854cf81bc272480e9fab2690eb48a6fb5190ba2b1deb8c21a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/codex-inline-judge.json", "bytes": 2792, "sha256": "b219f10014aa38224984aaf639e0aadfe16121aedec0bca2309c83c85fbd236c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-20T08-59-10-971Z-do67/postmortem.md", "bytes": 9631, "sha256": "572fe8ccbea7ba254474426fcf9f973defe301f11089c50f4fa28fb6b5914cb8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/f6fff9016b.json", "bytes": 3618, "sha256": "cf43810cb1eb15d2b53dbf0630344a62d18a3bbf6297d195bd7d0269d02a4cbc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/f6fff9016b-desktop.png", "bytes": 11652, "sha256": "1ff479d63a462e236beaf0634bccd56fbef6f5ed4975ddcf820cf6821ef1f375", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/f6fff9016b-mobile.png", "bytes": 9177, "sha256": "9f2ad397ef7066ed603a7eb094ca0ca5a13945e61441d84e51091b3ce1843f0f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/workspace/tank-combat-arcade/index.html", "bytes": 6282, "sha256": "b05999a132bd78030ad216811bc8375224ffae10b1c45d58d40fd935d01347d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/result.json", "bytes": 911, "sha256": "7e666ed05a0655fbef2f4494c4bb616dccb9b2c3c5637259ac3bf70786fbf551", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/score.json", "bytes": 3196, "sha256": "99ed3ea342a3497799e3d27f60bb822ae4b7398173b0e8926490019046f162f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/metrics.json", "bytes": 2373, "sha256": "49a3417387872da8934da1b49172366c841c50f8cb5f71851b13a5502c2245ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/facts.json", "bytes": 13065, "sha256": "138fb26cc4a048c21aadeceaa20b69395e3372ce00a139a8426e715ed16fe914", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/codex-inline-judge.json", "bytes": 2785, "sha256": "350c0272f4e5d2e1897127a76ad0fe3e18c5ba83c0b3d180b867f595dd56fccc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-20T09-32-40-241Z-lng2/postmortem.md", "bytes": 7573, "sha256": "ae53624cfd78ce7cfb5a4313f6d9a0c49bd693fb38a3d1ad3d7ed1e32d275007", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/525d991a7f.json", "bytes": 2159, "sha256": "d6de18f2ca9e93ff4db655f217cb8c22d413d95738208914de9bcec4e26478e6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/525d991a7f-desktop.png", "bytes": 10192, "sha256": "dfeeef08062894c5b55ae8a26cbfcb985ebd0497aaee1cc0f70798e6ea3db75b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/525d991a7f-mobile.png", "bytes": 5018, "sha256": "e297b3024df22925c77b770029f9d925022bae5518e6760f2ec85083950ba0a2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/workspace/tank-combat-arcade/index.html", "bytes": 7816, "sha256": "3ffd9ddc4f16a08d5f3728fc14a126093e152b21b5afae147dc5c2babfea0e55", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/result.json", "bytes": 911, "sha256": "99e5e556b0f5803036feb7d4d526bdab6057c3210ad9496c8cd44a5c7b598e81", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/score.json", "bytes": 3197, "sha256": "5a77dbfd6d955bf5b3b7c53f17f9bfea91fe88dd444d97ac742283838f93c6e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/metrics.json", "bytes": 2341, "sha256": "9b2361d64d1bbd4710aed757b7a3b49c492a33360ca4d47ea4bc98fb8e3dddb7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/facts.json", "bytes": 13429, "sha256": "0f835dc5e030df5857211de21c8312351478ff86107556f018cfd109301587a7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/codex-inline-judge.json", "bytes": 2909, "sha256": "5b397eb7f0c50ab44fdb479579705bc536a85523cfe617e6b9da0d96e717d123", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-20T09-35-30-484Z-z8ak/postmortem.md", "bytes": 7603, "sha256": "1e96db9ecf55c8e28db211155318ad6042a29d313c24f1bc139e09be6c77751e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/22cda23313.json", "bytes": 2358, "sha256": "e34e71b31376aec9d26d49df1a3f18684109153b23b62c56a0ac03b256e794a6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/22cda23313-desktop.png", "bytes": 8000, "sha256": "3a7cfd97a435ece9d03114197c7ea3cf4972feb4d3ec927a3805631d8c005c74", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-browser-evidence/22cda23313-mobile.png", "bytes": 5475, "sha256": "4848f6b0fdd8fd15490c63aa17ba0282dd2e84f5d7b0df22717b084933b48f6a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/types.ts", "bytes": 421, "sha256": "f06953b45de38be3ee557f495a8e95310eaccf49bf5b4654306d8b44039bb0e5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/store.ts", "bytes": 803, "sha256": "d674b67c353f985984784d2d03622b0576513988c6596e3cd0fab0a7678b1906", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/handlers.ts", "bytes": 652, "sha256": "f35af87ff55dc7c41d2d113e34eb2784d70818d7cfd68b97861e570850f050e7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/migrate.ts", "bytes": 844, "sha256": "54bea66ee57be7dc7f3c4230e6461b7187e47a347be4a75c561b066d8e2d3e2a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/tests/migrate.test.ts", "bytes": 1657, "sha256": "3a092472068f17aeebaa1a54aab65beef82920e3a112e044b1bc682ac05b4e35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/MIGRATION.md", "bytes": 2924, "sha256": "1d3b47cf7995899b99fb5087328ccab339b53f41a5f00fe25e716b617d34c9f6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/types.ts", "bytes": 421, "sha256": "f06953b45de38be3ee557f495a8e95310eaccf49bf5b4654306d8b44039bb0e5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/store.ts", "bytes": 803, "sha256": "d674b67c353f985984784d2d03622b0576513988c6596e3cd0fab0a7678b1906", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/handlers.ts", "bytes": 652, "sha256": "f35af87ff55dc7c41d2d113e34eb2784d70818d7cfd68b97861e570850f050e7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/src/migrate.ts", "bytes": 844, "sha256": "54bea66ee57be7dc7f3c4230e6461b7187e47a347be4a75c561b066d8e2d3e2a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/tests/migrate.test.ts", "bytes": 1657, "sha256": "3a092472068f17aeebaa1a54aab65beef82920e3a112e044b1bc682ac05b4e35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/workspace/user-store/MIGRATION.md", "bytes": 2924, "sha256": "1d3b47cf7995899b99fb5087328ccab339b53f41a5f00fe25e716b617d34c9f6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/result.json", "bytes": 909, "sha256": "06b63283ed8970b5195367628128ccc67e560241cb9956a6420bad4b87763aef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/score.json", "bytes": 2950, "sha256": "9bb06efb96cb0c171c13e620f10ff6e3c36686c60cc760f349887940d0dd0aca", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/metrics.json", "bytes": 2325, "sha256": "2e43a0f62ca51f4eb791379fd89352fc856171c9cf98bf0dbea18f27d78351e8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/facts.json", "bytes": 14404, "sha256": "ca28c7caeb1cb94755998e9619e4f48692b8e4bbc403ee4b38d233df03bc6835", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/codex-inline-judge.json", "bytes": 5556, "sha256": "c2af20bbfaa1bb4feaa582054920d8a49d5262e69819a6f2b672eaad0ac8da72", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-20T09-58-55-369Z-vl56/postmortem.md", "bytes": 7517, "sha256": "dbb86f307d2cec1ab41262a4b794865dd4c9ab08c1de7889dc2220ebbec8b453", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/09fc1b7dfc.json", "bytes": 7733, "sha256": "ffedad1f9c0112760eadd6f2df44a7ee9c4dba769f23a28d445f33146713dfd9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "32ed67b8728ed6a20fd6434d95ee93e16921b68685d98cca60465f3b43414bd6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/handlers.ts", "bytes": 619, "sha256": "75e6ed60c776cb1b0cbfc803ae8c4c6e7f51910209ae90f727cb5644e5820553", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/migrate.ts", "bytes": 354, "sha256": "2df3db7e81cc0d5063c2b11a5e87eeb5d631d7f80246c28525b5691e07cdf467", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/default/package.json", "bytes": 128, "sha256": "6aeee50e1df303e1c6f34302b42f08639395284c0650a376550390aa012b9727", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/default/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "32ed67b8728ed6a20fd6434d95ee93e16921b68685d98cca60465f3b43414bd6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/handlers.ts", "bytes": 619, "sha256": "75e6ed60c776cb1b0cbfc803ae8c4c6e7f51910209ae90f727cb5644e5820553", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/workspace/user-store/src/migrate.ts", "bytes": 354, "sha256": "2df3db7e81cc0d5063c2b11a5e87eeb5d631d7f80246c28525b5691e07cdf467", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/result.json", "bytes": 1622, "sha256": "e2e8d1964bcdceb22d5dd19352a6c0ac39911f48f2f317388ae57e033ed68864", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/score.json", "bytes": 4545, "sha256": "22bce675fc77bfcc322253b6e73bf79fcbb1183a249ee8826d9afb7c3e7e3f98", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/metrics.json", "bytes": 2372, "sha256": "b81370add7b356dfaadc4f5ccb85fd4f59ead9484351fed6f693f2ff35ecd225", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/facts.json", "bytes": 16577, "sha256": "dac604ee656f3efb8083a87d3e9f994326fcc85a83a7ca288f0c3f7d5f3c116d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/codex-inline-judge.json", "bytes": 5477, "sha256": "3239cc12ccabb798b67e19760ed74994f3618d5205e5a2afd64bd1b454ace461", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-20T10-01-58-015Z-tjuj/postmortem.md", "bytes": 9235, "sha256": "5f389a870c84a0f5e50dce8ef303542dff638cf6529b1fa04a16fe3f24b6aafb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/ced997af06.json", "bytes": 7010, "sha256": "97d061e96a51e469144f8957f8989c2728b2c682ce7710517b9829cc41b46532", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/types.ts", "bytes": 362, "sha256": "ab0c7c1964e365a0228f598ae347d40e8a0880aeafa4435ae5477dda5a38e69e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/store.ts", "bytes": 845, "sha256": "a6e48d3d9c28ea06b3316622ce19f5d8f18e1c24e0f8008518d3dc367fd0c816", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/handlers.ts", "bytes": 658, "sha256": "813604532a20ee91a46d36755be626cda3ea2b53ac05f019a0068f15dbd5b2a9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/migrate.ts", "bytes": 452, "sha256": "8463f4e06f92776f6912f17ff7b606ab79422ea83c91c7156ba212485f683c81", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/tests/migrate.test.ts", "bytes": 1251, "sha256": "02ea7edf2e1dc27cd61a89e64e4f4b68dd1abdabfa3fe13e639c6b506f44f2e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/MIGRATION.md", "bytes": 3482, "sha256": "e1a8ba95d31dab96a00627df36fa559fbae080fbe72c2ab42e2f11bf93e16f35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/types.ts", "bytes": 362, "sha256": "ab0c7c1964e365a0228f598ae347d40e8a0880aeafa4435ae5477dda5a38e69e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/store.ts", "bytes": 845, "sha256": "a6e48d3d9c28ea06b3316622ce19f5d8f18e1c24e0f8008518d3dc367fd0c816", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/handlers.ts", "bytes": 658, "sha256": "813604532a20ee91a46d36755be626cda3ea2b53ac05f019a0068f15dbd5b2a9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/src/migrate.ts", "bytes": 452, "sha256": "8463f4e06f92776f6912f17ff7b606ab79422ea83c91c7156ba212485f683c81", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/tests/migrate.test.ts", "bytes": 1251, "sha256": "02ea7edf2e1dc27cd61a89e64e4f4b68dd1abdabfa3fe13e639c6b506f44f2e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/workspace/user-store/MIGRATION.md", "bytes": 3482, "sha256": "e1a8ba95d31dab96a00627df36fa559fbae080fbe72c2ab42e2f11bf93e16f35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/result.json", "bytes": 970, "sha256": "19f2b9a9fb8c77ec02617562b3bb464d82eaa32c5ac599436690e4241c73a408", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/score.json", "bytes": 2973, "sha256": "6acf049556ccc459e3065357976522ffc8c635cd18760a1f116be08c019b8a50", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/metrics.json", "bytes": 2406, "sha256": "c1a42eb9874e07bec06e1d34fbcbf3137a13f85fa470e302bc77e251fe64a784", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/facts.json", "bytes": 14879, "sha256": "0ae2f6d1904a3961e89ab22deb192c7bf3ea6e5ec785d0e0028802a1982b89f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/codex-inline-judge.json", "bytes": 5859, "sha256": "3f1ee37a4eb6cbf1cda647fc44396c3604cd341d9359fca5165a1d078a95e716", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-20T10-04-13-360Z-69my/postmortem.md", "bytes": 7530, "sha256": "03749dad754d5d7f5042a5b13a08c23766f3ea682198c82edee27786a30e5834", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/739839695e.json", "bytes": 8159, "sha256": "35c4dc79b41e8527cda67eb13cc1a9da35fdad3df5a7c086137377d116dd0641", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/types.ts", "bytes": 370, "sha256": "64cae9473d274f167d6dcbc09c75f1345f4881caded992a6d33f9ce4ea662eda", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/store.ts", "bytes": 734, "sha256": "241c5258f2bf712c79b27b5871f8b44fd0581b753a3e3d63dd3038541c81203a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/handlers.ts", "bytes": 537, "sha256": "f65fda80a1706958d14b1f14a3ddd6fe9c639d1aab38eb7b184277ae60c5d56b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/migrate.ts", "bytes": 878, "sha256": "4c8f05c4f573e5a1d668ada7e0b6b8ea5ce362c7c687d2446780ee2c0c6bc4aa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/tests/migrate.test.ts", "bytes": 1138, "sha256": "3025c1dde8c3134807d6064e071ab2c821e52ba52451eb4933377eb4c5147e56", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/MIGRATION.md", "bytes": 3360, "sha256": "0461f56869e5e2ef6e4bc7250ba9e54f8288c094791c9d83d5acee3a02600644", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/types.ts", "bytes": 370, "sha256": "64cae9473d274f167d6dcbc09c75f1345f4881caded992a6d33f9ce4ea662eda", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/store.ts", "bytes": 734, "sha256": "241c5258f2bf712c79b27b5871f8b44fd0581b753a3e3d63dd3038541c81203a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/handlers.ts", "bytes": 537, "sha256": "f65fda80a1706958d14b1f14a3ddd6fe9c639d1aab38eb7b184277ae60c5d56b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/src/migrate.ts", "bytes": 878, "sha256": "4c8f05c4f573e5a1d668ada7e0b6b8ea5ce362c7c687d2446780ee2c0c6bc4aa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/tests/migrate.test.ts", "bytes": 1138, "sha256": "3025c1dde8c3134807d6064e071ab2c821e52ba52451eb4933377eb4c5147e56", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/workspace/user-store/MIGRATION.md", "bytes": 3360, "sha256": "0461f56869e5e2ef6e4bc7250ba9e54f8288c094791c9d83d5acee3a02600644", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/result.json", "bytes": 905, "sha256": "e82f927edb279a9cbe414e61749b22b4001c3cf4e6391272b3866561f1c8ccbe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/score.json", "bytes": 2947, "sha256": "512e24459074eb4ecc9a097c5027d5a805c752250a5a2a9aa93176b1e6dd3895", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/metrics.json", "bytes": 2337, "sha256": "a8a2ed2c273ccd506a5eea1f8c1f2d84d33aec650ce478f8002b23a643b490c8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/facts.json", "bytes": 14697, "sha256": "ca90467ab9577f9c716fa5653a67a8a7aac96f16362491245898c48e819230a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/codex-inline-judge.json", "bytes": 5516, "sha256": "0750aabbd3d36e3029477c6e43112de5ace489342d02610f8d58306e164aaf94", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-20T10-11-52-214Z-aeu0/postmortem.md", "bytes": 7536, "sha256": "6621c63a87b7302a026f0709a26207433821a61d576f886543cc549722a7f35f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/4ba2ec696d.json", "bytes": 8448, "sha256": "955aff853640fbde7f2ab68d6322ea51d0cb2fc4368a85389555776b3ee7ca58", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/store.ts", "bytes": 908, "sha256": "0955f63b59c86d5112fccb046c79770e14ad91f394c114abc9200b7d1c89463f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/migrate.ts", "bytes": 635, "sha256": "ac94f4815f6a3e88d835c9df72cd4138020b8084b0cc8cc0d68b123168ad7e02", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/tests/migrate.test.ts", "bytes": 1060, "sha256": "814c1868cb45cf98d71914730d87c393ded9247446bd38a24a5edf4ff748bc48", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/MIGRATION.md", "bytes": 1514, "sha256": "0e202c94cab92e7c1b478674a8d47f48c9f0875d1d40c0a32c8eafd90618fb96", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/store.ts", "bytes": 908, "sha256": "0955f63b59c86d5112fccb046c79770e14ad91f394c114abc9200b7d1c89463f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/src/migrate.ts", "bytes": 635, "sha256": "ac94f4815f6a3e88d835c9df72cd4138020b8084b0cc8cc0d68b123168ad7e02", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/tests/migrate.test.ts", "bytes": 1060, "sha256": "814c1868cb45cf98d71914730d87c393ded9247446bd38a24a5edf4ff748bc48", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/workspace/user-store/MIGRATION.md", "bytes": 1514, "sha256": "0e202c94cab92e7c1b478674a8d47f48c9f0875d1d40c0a32c8eafd90618fb96", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/result.json", "bytes": 900, "sha256": "fc4e13e77bb4ef43b1a171eba070fdd30b89909eb3cb8aea564deef665af2ec3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/score.json", "bytes": 2933, "sha256": "628c836b4a6cf092a96d7bdace061622135f76af04a19aedf9b61f71997a6c86", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/metrics.json", "bytes": 2379, "sha256": "3cfb2cacab21d03eca18d6a26f6c5cfc40175ee3ff41ad16072266280111550b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/facts.json", "bytes": 15605, "sha256": "fbcbb129d6e378100660a6f900f296ffa98759ae9443962876fac65592ac4101", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/codex-inline-judge.json", "bytes": 5457, "sha256": "76404000f85b707a551ad4165d36a463656fa8a6a0d6b13114d960cf88ce7b23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-20T10-23-10-457Z-w6v9/postmortem.md", "bytes": 7468, "sha256": "b4d55d75aae134aa351403a101c4cb9fe85693c0fd2ca1d866e2b655d718cc68", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/56b4f1ce18.json", "bytes": 7753, "sha256": "598110b7f268b7b4631cc0b9e821996a3dc52e0daa8b7dd9e3f0a6b00c4fd4ce", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/types.ts", "bytes": 193, "sha256": "0c9a266194352ea16117ff3b7dc5471f7548d9eca2b1f6dbe5c9d7564ba136ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/store.ts", "bytes": 916, "sha256": "50013f5077497552b29e7a64ebc6a9682b1fdcff3d1546f85d5e0c5d677240a9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/migrate.ts", "bytes": 1174, "sha256": "0ff62f26ef4988022235ce39c1025528f3e19b843d4ef661976c10f6953c28a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/types.ts", "bytes": 193, "sha256": "0c9a266194352ea16117ff3b7dc5471f7548d9eca2b1f6dbe5c9d7564ba136ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/store.ts", "bytes": 916, "sha256": "50013f5077497552b29e7a64ebc6a9682b1fdcff3d1546f85d5e0c5d677240a9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/workspace/user-store/src/migrate.ts", "bytes": 1174, "sha256": "0ff62f26ef4988022235ce39c1025528f3e19b843d4ef661976c10f6953c28a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/result.json", "bytes": 1534, "sha256": "bb4a9fc16974031de02119fd3cf50225e7ec5f1bc16b443974ffe31309f50544", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/score.json", "bytes": 4430, "sha256": "d5ecad0adf04e4195ab4d2c4386c6502f0317b3870bb1c056aec33bf6092bd0e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/metrics.json", "bytes": 2342, "sha256": "573acccfd6f391a5a22f91e14cf7841224eb2330c74aaadf384894981229462b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/facts.json", "bytes": 16930, "sha256": "c4204cb7e32c3f91bfbbf2f2e5322250e24b8726a9381a1d1689c60504ec363c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/codex-inline-judge.json", "bytes": 4354, "sha256": "da4537dfab21d448891d9d2e4d09a95836741708bda0ee9deeaea6c82c2d576e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-20T10-26-49-419Z-i2x7/postmortem.md", "bytes": 9348, "sha256": "ad472933aed37b9ccb2b9b477103efd3831db3f4275fee2688909c91f659a9c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/1cfdb5b3b4.json", "bytes": 7194, "sha256": "f9299fe02d98251f5f5d7fa0113433cf35d9466eb379128f821319309c96b6ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/types.ts", "bytes": 341, "sha256": "5ebfd444364649122142ce621a072ae5b218cc4bc2110936679ad4e26e0657ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/store.ts", "bytes": 1088, "sha256": "c8703cee25f73689571e8e104dff4de99539c1a1b2a7bccaf1fd94e485edc9fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/migrate.ts", "bytes": 279, "sha256": "798873d4a1dcd91c9c1f6584b893b4a5eea2fb7f37010fa9009c19da92d611b6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/tests/migrate.test.ts", "bytes": 1222, "sha256": "70b38acedc7c9faebcb356a3e10c3b603f02840bad115351b520f98d95dbdb38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/MIGRATION.md", "bytes": 2658, "sha256": "ff68529f6499201f0bae98481542cd9e029fc481ddd2314cb0f79f0b8c5a3b7c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/types.ts", "bytes": 341, "sha256": "5ebfd444364649122142ce621a072ae5b218cc4bc2110936679ad4e26e0657ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/store.ts", "bytes": 1088, "sha256": "c8703cee25f73689571e8e104dff4de99539c1a1b2a7bccaf1fd94e485edc9fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/src/migrate.ts", "bytes": 279, "sha256": "798873d4a1dcd91c9c1f6584b893b4a5eea2fb7f37010fa9009c19da92d611b6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/tests/migrate.test.ts", "bytes": 1222, "sha256": "70b38acedc7c9faebcb356a3e10c3b603f02840bad115351b520f98d95dbdb38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/workspace/user-store/MIGRATION.md", "bytes": 2658, "sha256": "ff68529f6499201f0bae98481542cd9e029fc481ddd2314cb0f79f0b8c5a3b7c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/result.json", "bytes": 899, "sha256": "edcacdc176fb087320739efc97562243cefc8470fd9136fd009479c73e127671", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/score.json", "bytes": 2921, "sha256": "4bf3bc5ebb37d2d76566c95b26b60af56588b060f1aa9de9e66e80f6d655c2d8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/metrics.json", "bytes": 2382, "sha256": "b4e9d47c885ae476402dd8f33eaa8a980da2ec1e4ec57361025149cb0fdc1ded", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/facts.json", "bytes": 17858, "sha256": "191c5fb3f0d4d7dfa754dbb2f045872710f389bc4a5a4565311646d8c5d81ab1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/codex-inline-judge.json", "bytes": 5475, "sha256": "75a6a7bf76ea6ec254599d750098f1276afad3a7081d48fb7016da380168e07e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-20T10-29-50-919Z-9o10/postmortem.md", "bytes": 7471, "sha256": "4a92f6dedaea1e4bda345d53d63dfe3e8c23b4a355caa4db9c9da170887966b6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/695b7a235c.json", "bytes": 7754, "sha256": "99c671c64af8acc21cc3998c8e329bb82c340342d704a34697ce227cc8603991", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/workspace/machine-kata/src/machine.ts", "bytes": 2305, "sha256": "5199fb0faae9d4a901d8013b1bbd96e12ff22feb766e2d6e3de8f44188692561", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/workspace/machine-kata/src/machine.ts", "bytes": 2305, "sha256": "5199fb0faae9d4a901d8013b1bbd96e12ff22feb766e2d6e3de8f44188692561", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/result.json", "bytes": 2112, "sha256": "f3468f68983e9fede6ba75c373d6da8fd10deb63eb2ce3f0f0bcf79a3482a3e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/score.json", "bytes": 5295, "sha256": "3732e9428f43c734a43533cb764a57ebd23b6725b5e281f41cc5e2603e979ee7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/metrics.json", "bytes": 2343, "sha256": "94596a7c90be1d96047f6824596deb6b018e388dce4c7f02bf525eb050238b3a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/facts.json", "bytes": 20254, "sha256": "ccd1d095aeda1ed5ba4c688d6f56cdfd0a8d8c2e593dec3b3dd7eb30bbf78acd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/codex-inline-judge.json", "bytes": 2686, "sha256": "2d0f195cf7d3712b498c8977bd642ee6b63ff00a116027e7feca8c5742c6ca3b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-20T10-33-45-802Z-8cnw/postmortem.md", "bytes": 9323, "sha256": "1e884dcdc7a45fe9459f0952a2961ef975d10a96add54d76e23f5e89292afda7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/23265c1535.json", "bytes": 46963, "sha256": "ef10ca0c59fd33db115435c16aaa9bacfb1c84b35115c306b6115486d843152b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/workspace/machine-kata/src/machine.ts", "bytes": 1850, "sha256": "d5612e67fbbb709e484f0d9f79c8263edb6b7b00cf7ca1c3d6b15d630a3c24d4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/workspace/machine-kata/src/machine.ts", "bytes": 1850, "sha256": "d5612e67fbbb709e484f0d9f79c8263edb6b7b00cf7ca1c3d6b15d630a3c24d4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/result.json", "bytes": 2591, "sha256": "39ecef31fa16696d85d6e51f1f2661efc5dfc253b6407b2a3161e2b10ee0d708", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/score.json", "bytes": 6074, "sha256": "e58cb40065b6dc23d7141120e7d2514c6d164857aa3a2b964488565f3df68f69", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/metrics.json", "bytes": 2384, "sha256": "f85b6ade8a81f4da048f147c79789f086e0cd8ae13e076c9372cfab9b5d667d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/facts.json", "bytes": 25263, "sha256": "10cf8f4a35dd19129a684e4b95b1f5d7ae18f96525a277b229a6499c532315db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/codex-inline-judge.json", "bytes": 2727, "sha256": "28f1f7512ad94a2def70cbb76b1c97b23f5716d222dbb17a191611b1b2b12826", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-20T10-41-10-176Z-na52/postmortem.md", "bytes": 9459, "sha256": "a60267169a5d975a0a292afb36df6828be6f4904faea31a4179f76d932390f2b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/3abdb0ce27.json", "bytes": 28701, "sha256": "c37d8100c49cd71a35c06247cda281a57647c6fae1cf9593c99e876027bb6989", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/workspace/machine-kata/src/machine.ts", "bytes": 1594, "sha256": "365777afecc0ff037c08e99fbff5204c728c1adf81b07379cc8bf387fb73f970", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/workspace/machine-kata/src/machine.ts", "bytes": 1594, "sha256": "365777afecc0ff037c08e99fbff5204c728c1adf81b07379cc8bf387fb73f970", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/result.json", "bytes": 960, "sha256": "aba81014e9769771e41836e72a83c38a2da0124fc5aae53a89c4e6214dfd38c5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/score.json", "bytes": 2879, "sha256": "471176008a658e36118cf2a68548a187b8407d0cf091f07777b1dc98a261c3ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/metrics.json", "bytes": 2353, "sha256": "e6fc81254cf7405fd0490c99d55b6cc812da14c2238ba7756181bfd56febff10", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/facts.json", "bytes": 13170, "sha256": "f480cb6182bf0efe1d13787584e8fea4e147515ebc74245c0c025d4d3b339831", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/codex-inline-judge.json", "bytes": 2793, "sha256": "d57c2a173d54d99a6b32d933bb261569311ee114e063a7d27c227b00b45294b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-20T10-45-32-018Z-8tu8/postmortem.md", "bytes": 7472, "sha256": "39df9031a04514d0548974c6034d7b597eecab2ddf88c64a3033101b926f2364", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/e7bf09d515.json", "bytes": 43461, "sha256": "5b3c438437a460292ee2a11677af82017c9c521d7b6d50d33c3c8eff7c0af2c5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/workspace/machine-kata/src/machine.ts", "bytes": 1912, "sha256": "79f2dcb6928065a0c754359fdabe52d56a59008ae179c9525548f95c009369bb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/workspace/machine-kata/src/machine.ts", "bytes": 1912, "sha256": "79f2dcb6928065a0c754359fdabe52d56a59008ae179c9525548f95c009369bb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/result.json", "bytes": 895, "sha256": "0d7060a33693af71d8eab29d0f4c65dd67a83b6ac0eabc553e10b917104c12e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/score.json", "bytes": 2855, "sha256": "39afa7e87693e5ee2bd07438916f03e0cb092364c5837bf73dd8baafc928841d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/metrics.json", "bytes": 2317, "sha256": "d477515c1fc031191d1ae7691335692f953d65de5b83e4f28f9f67dcc7455f42", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/facts.json", "bytes": 12579, "sha256": "900377b4709e9e47269e18c158435fdb4f7e95a761785fcdbb3ba7dd39d94c29", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/codex-inline-judge.json", "bytes": 2711, "sha256": "10978c8311ebf080bdd4ce02ceefe3ccf0f6bf91986ebaa13d519007e5a2de85", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-20T10-48-57-096Z-odpb/postmortem.md", "bytes": 7483, "sha256": "39a0489311c614d99bc09ad546be11c872f4495e58d993825c2fd9c7dd93c7b5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/e6c6ed3b40.json", "bytes": 42446, "sha256": "2f7290b7312ad7858844eb208703532793e5cc0310865c572e85d952166fa7e9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/workspace/machine-kata/src/machine.ts", "bytes": 2185, "sha256": "85e7cd2c87b3fdd1b1e22edd76e856265eb5ca864a58b62196ab0a91c4f19b34", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/workspace/machine-kata/src/machine.ts", "bytes": 2185, "sha256": "85e7cd2c87b3fdd1b1e22edd76e856265eb5ca864a58b62196ab0a91c4f19b34", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/result.json", "bytes": 889, "sha256": "0c9aedb82fca488eb614e3d0c8f66e924ca9f169c7e28750c9b3c39598068346", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/score.json", "bytes": 2851, "sha256": "c10afc2d12497d2d6e64c45c68d4fe8fd11c6c6076e78247eb70c552db304ddd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/metrics.json", "bytes": 2319, "sha256": "da175e035b0f3ea2738fbffea84229627222892d8a8e6aaea1a906f9d461738c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/facts.json", "bytes": 11971, "sha256": "08fb6861f62425be14187837dc51fd27fbbc993a7151f9ec379bbddc05f26751", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/codex-inline-judge.json", "bytes": 2675, "sha256": "4406b3906e1e27e4cf047a998407ea551573f7df9d9caacd36b090a257c8b034", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-20T10-53-31-609Z-rxgq/postmortem.md", "bytes": 7451, "sha256": "686bdf7574678a17ba8e156e64ad28819e33275740ace57849e731c951f36a73", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/f52dd22705.json", "bytes": 42439, "sha256": "7a9ef458bb47724df5d3cb7f8df898fd5071c05dfde4672297af38c33a9806fd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/workspace/machine-kata/src/machine.ts", "bytes": 4011, "sha256": "543bb1ca5226df7403957e19cd70d146e656835dd04fc564b0f5c5eef64dda4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/workspace/machine-kata/src/machine.ts", "bytes": 4011, "sha256": "543bb1ca5226df7403957e19cd70d146e656835dd04fc564b0f5c5eef64dda4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/result.json", "bytes": 888, "sha256": "919596e0602ad310d98f65f60906ad8677beeeb648ee848568b0143a2fcc03d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/score.json", "bytes": 2838, "sha256": "f539883a84ad54987edf80ee312bc9b8527e54d08b12d425b185fb1b69adbc13", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/metrics.json", "bytes": 2323, "sha256": "1a27dcd6ea61086018ebc9eba15402907b2180b7e5cd01841af93f96c886c7ce", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/facts.json", "bytes": 12066, "sha256": "dd24485f2f0d36b69a63713343d8e9cb6775f2228972f0c74eb515742ea819cb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/codex-inline-judge.json", "bytes": 2676, "sha256": "7925d0e0c4fbd6a734af0bb678458238bc8a12160fd4eb2530f14cf6d5c1d569", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-20T10-55-13-256Z-5zcv/postmortem.md", "bytes": 7415, "sha256": "1a4f4c2c189377169920550f74e85ac05cbb5ec551c74472f30e8b6ce4b4dda8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/c237b146ec.json", "bytes": 42439, "sha256": "1c9107c14a1e1c53f553b5fbc363a916078f07ba1c1434278b9bbad1eeabb5af", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/workspace/machine-kata/src/machine.ts", "bytes": 1652, "sha256": "e2c3af3f22f900bcdf05fe0c3eb4d7410bf57c0598ba451f2942bcda453ec1e0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/workspace/machine-kata/src/machine.ts", "bytes": 1652, "sha256": "e2c3af3f22f900bcdf05fe0c3eb4d7410bf57c0598ba451f2942bcda453ec1e0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/result.json", "bytes": 1738, "sha256": "62f12075c9279f79904216028d3b4e7481d9b68acd459e7c1c31d128b03aaf47", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/score.json", "bytes": 4844, "sha256": "9415160e64c6246880cacd9ebc332e3ab75c85984ea7934318c558e7fca491bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/metrics.json", "bytes": 2336, "sha256": "d163a215ca7548b01870cedbb6e3a03bf8f4a073833d82f94d3e02ab8e540110", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/facts.json", "bytes": 17049, "sha256": "0c178150b79c2b1b56842b318037d964d0be67c570854fd0fc130ef8329c1b62", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/codex-inline-judge.json", "bytes": 2688, "sha256": "d8c9653051c418f861eb5aefee8a7559ab5dae51004793c615464b9713d8bf70", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-20T10-56-45-838Z-0wcm/postmortem.md", "bytes": 9455, "sha256": "6ccddecab6e0c02b4c1810d8f8754032af4898338adb739b8d05b1d03c3996d5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/9e1ae91c78.json", "bytes": 42379, "sha256": "08a8b4ab4c98842039222280b47a1fec1cb0fe4114c0c1321f182f2359dfd795", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/result.json", "bytes": 849, "sha256": "7906d9ef3767fbf81fb303d52ff9e5fa24ec70a3615422f6251a7cb83f5acadd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/score.json", "bytes": 2797, "sha256": "1e0ff9f535b87f315bd109c04a7ac63db0c411324a3fd5f302e92afee99991e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/metrics.json", "bytes": 1306, "sha256": "eb63be92964c4c6d4bc55351a29359a6184f89eed08674e56f4cf5661a4c4b22", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/facts.json", "bytes": 12018, "sha256": "4437b83ef81ed2605d72fa256796dcd93a124660cafccc97bbd500532bd2a569", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/codex-inline-judge.json", "bytes": 2650, "sha256": "dbfb2506e084dcbf0a69a19e54f0168c3cfe9d9e4ad3c123e97ac9ab46b285a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-20T11-17-50-508Z-mkyw/postmortem.md", "bytes": 7374, "sha256": "839fcc25efbb4c774fdf2c0affc60d70f826aaee834b6b7a048f7f5771f706a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/a4a2677022.json", "bytes": 8573, "sha256": "0e2e834905514f4ad86097467316562d6a0213751c988f963ba825446e7700b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/workspace/pagination-service/lib/paginate.mjs", "bytes": 1209, "sha256": "0e99975ec38533f40d28d03a6137d978e8199bbd806b9acf801f9de5cdbefc42", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/workspace/pagination-service/lib/paginate.mjs", "bytes": 1209, "sha256": "0e99975ec38533f40d28d03a6137d978e8199bbd806b9acf801f9de5cdbefc42", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/result.json", "bytes": 922, "sha256": "2c6feb44871ee4c6b0f025f7270f9b96e81d1fcb8edc16430918de3d679b3017", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/score.json", "bytes": 2838, "sha256": "093e8aac8c3430f0bb5e696082e710b5050fd8dbfa9fbf684626f9dfeed60dd3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/metrics.json", "bytes": 1311, "sha256": "095d59109bf8ca6393051e7e92b34292e466d1cbb4fcb8335315c29e275f93a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/facts.json", "bytes": 10697, "sha256": "e5a71f319cf75d0e12ee54ebeb27167d694c0f9c1bff704267501b5676a64dea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/codex-inline-judge.json", "bytes": 2722, "sha256": "ad8208c4e2147c0d50ce1b3b43ceb3d40fe5b2c1a7eb05474b3f330aa5bb8019", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-20T11-20-03-090Z-kg51/postmortem.md", "bytes": 7396, "sha256": "21a446f23f2587ff31f62a1fa1c3069d8fd33672af46415d9228bbaa71e153b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/3a4f70bbe9.json", "bytes": 8668, "sha256": "4cee31eb63a455fa4f7707839729c8fb771d247fc59ab3badc42ffda157165e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/workspace/pagination-service/lib/paginate.mjs", "bytes": 1218, "sha256": "c950d0989a816e70b007a851e1c56404ed0f470e37676a246fdf809affaf9282", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/workspace/pagination-service/lib/paginate.mjs", "bytes": 1218, "sha256": "c950d0989a816e70b007a851e1c56404ed0f470e37676a246fdf809affaf9282", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/result.json", "bytes": 857, "sha256": "a98b8957ea988b2076c37267d5e870ccb577f5af1ce6b488e2ce9c05679bd8e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/score.json", "bytes": 2812, "sha256": "b76e80d4b64146e35566a052c1a540f7be6b5e4f79653671e01ba80df73a0eaa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/metrics.json", "bytes": 1307, "sha256": "5fb8faed699926cb882fca1296f1bbf667026557036f15020ebac5956f14b705", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/facts.json", "bytes": 10557, "sha256": "d51f7fb4c12c19ff0df0ee2355723cce0761b0ff3c369cbec7255a38a9535812", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/codex-inline-judge.json", "bytes": 2622, "sha256": "cdcacc19f4963e913bb500d2c403b250980c1b69de984840ebc9495ce8d1667f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-20T11-23-22-605Z-iv23/postmortem.md", "bytes": 7385, "sha256": "0635506d6896b786734a83ee98b6d21dcc2d2ca4c58568fa849c8983682bb103", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/a4d75ac48f.json", "bytes": 8551, "sha256": "b184c7ccf5c378d9b84d8e40787e33d3762316a257c24106ee23f995bc4b9280", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/result.json", "bytes": 852, "sha256": "aa18c13a6321e7489d8e743fb047e68f5377c9a6e50920f09372f19f687a6540", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/score.json", "bytes": 2816, "sha256": "7fefbd4bb68043cce8d0555fdb5e2c16716aea5f81c45b5d3827e2daec3af2e1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/metrics.json", "bytes": 1308, "sha256": "cddb238d3b353d3d5d2e8679cbb98b96f61a3d4cad4f6bb5020f76eef19ef35b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/facts.json", "bytes": 10913, "sha256": "a47a277ad0bcb7d3dcbbd79e35912d169e98d59d2748c1b967b243a64f23cd8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/codex-inline-judge.json", "bytes": 2646, "sha256": "97e83979132970a248049fec412b6fa318fc0a1a9319bb92e7fc59fa96d69996", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-20T11-28-51-093Z-xk6e/postmortem.md", "bytes": 7429, "sha256": "f19e92e39ee4eacb8ea742f8043579a8eb4f1d6d0f748ebfe37c26046d8e82a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/e3dc1045fd.json", "bytes": 8620, "sha256": "ced44b5a4098699e72b35b67f1c088586558805890ea2875d2622376507ab1a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/workspace/pagination-service/lib/paginate.mjs", "bytes": 1187, "sha256": "648c72b08714655c84925eedc78a430254ad3b2d64bde9cda8a2a35df300e43f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/workspace/pagination-service/lib/paginate.mjs", "bytes": 1187, "sha256": "648c72b08714655c84925eedc78a430254ad3b2d64bde9cda8a2a35df300e43f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/result.json", "bytes": 850, "sha256": "9724ad33f484db8b11e5d457183f54ba81fdd90bbf0e0b0f1c8f841b560e8c9f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/score.json", "bytes": 2808, "sha256": "50f584def7ada52a16e7b0871b80a39a9ee3e571e2da78a422c7e835f885f2df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/metrics.json", "bytes": 1304, "sha256": "7cbe0b89568ec77712cdc29d82d7ab25f3ddddc9eb48d3650bd93c340bf38233", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/facts.json", "bytes": 10422, "sha256": "4874d5968560186a7842c1b3e5ef382fbec98998df3585d4a1dd680ae56f5559", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/codex-inline-judge.json", "bytes": 2632, "sha256": "ee7d9610c3a246ac0d663e36112af9f3567224a43f0203e9369686711160abfd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-20T11-31-51-904Z-ysm8/postmortem.md", "bytes": 7398, "sha256": "9aeb9b1642c3a215e1c364d6939cf18ffda6d3d2db8d4ad49608b2c4550398d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/b43d6a3154.json", "bytes": 8540, "sha256": "f77dc669496e50fa37c1df663a8d7771fba1642455706aa88105e94ecfb5c81e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "41cd12dd72a6fd3e73c1bae4ea35e8241d4662e31dc0c0de0eadcbe4f01c4db5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "41cd12dd72a6fd3e73c1bae4ea35e8241d4662e31dc0c0de0eadcbe4f01c4db5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/result.json", "bytes": 850, "sha256": "432491712ceaf82bcfb01a74047616d49579c4fd720b84bf7c1dc98e41095b05", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/score.json", "bytes": 2807, "sha256": "3b8a7e383360b4a23d527c388768fcd78a02125ce507a47b58dfae1cdf3f202b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/metrics.json", "bytes": 1306, "sha256": "47d0733e65865e1f1bca5b40792901c0fc0dbb952aa9ccd5d3197f46ebaf2977", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/facts.json", "bytes": 10462, "sha256": "549730d691b1db8985802ca55584e1bf51bc02b6126c67e05b28459ca89fb346", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/codex-inline-judge.json", "bytes": 2621, "sha256": "f9705efbf3c0c66eace6ddaf4dfbbf7f9570444cf65fb4b9b4983a9c20b23258", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-20T11-33-23-460Z-4r3u/postmortem.md", "bytes": 7384, "sha256": "d73102ebd80eeadb5859e682770846bcdfbeae9cd569a0b96342468ff8838c27", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/f56cd2ab45.json", "bytes": 8575, "sha256": "f177c7718a7d6b62e305a27848efedd640761447db29846e402969c08d410c71", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/workspace/pagination-service/lib/paginate.mjs", "bytes": 1226, "sha256": "a90e859c98f5a93804ecdf94c0d696ac6fc26e45b9d662a4331046a16bf90b73", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/workspace/pagination-service/lib/paginate.mjs", "bytes": 1226, "sha256": "a90e859c98f5a93804ecdf94c0d696ac6fc26e45b9d662a4331046a16bf90b73", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/result.json", "bytes": 861, "sha256": "e5d0e37e31eee9090ca62d47383bfc94280c2c749ef04a656dde0a826bf2036b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/score.json", "bytes": 2803, "sha256": "79fa4c6101e3962e0b76091f8818fe6a5b3ea8850c6431937b572a600e883cc5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/metrics.json", "bytes": 1306, "sha256": "d651e5f5f7cefd18625bd54fe1bc0b6cc4435ef90263b72e1a910a619fd2c8b6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/facts.json", "bytes": 10694, "sha256": "ee5ac46c33f7ce971664467c2f5590a27d763481360cc63b67841eb6ee149a74", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/codex-inline-judge.json", "bytes": 2629, "sha256": "750d52de3cfecdd3b4268ddec14650f72d9acd0e8d990a1b3c5945a6e92d2254", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-20T11-54-44-141Z-3q3a/postmortem.md", "bytes": 7352, "sha256": "95fe411813e17c0452f77faffadec045e537b26e011d1b6ff66055006b5ee737", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-code-evidence/60812ef699.json", "bytes": 8591, "sha256": "6eba63d9a163ada1c68ae945933a097a110ae632065932ca74be971b66ba3a99", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "68c5c5800bca493ab34afcf02508284cdfc1f326ed0d2bbc8c93dfe129040511", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2263, "sha256": "f15a7c625669cf67727f99801d7f5584d7d369269e89da224ce2a7c99787a438", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "68c5c5800bca493ab34afcf02508284cdfc1f326ed0d2bbc8c93dfe129040511", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2263, "sha256": "f15a7c625669cf67727f99801d7f5584d7d369269e89da224ce2a7c99787a438", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/result.json", "bytes": 1308, "sha256": "79996d9a74941523b3792251830fe010d794793ed7147c22f4508730a5122a8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/score.json", "bytes": 4439, "sha256": "7b8f96e44942c364f9ee5e6f352ff2d77618b4698782324c8227b443533942de", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/metrics.json", "bytes": 1320, "sha256": "9f5aa323bf8406e4c7af2c9708d69f668436c59081f53502a9e8e94f999a49c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/facts.json", "bytes": 14412, "sha256": "77cedc96f6316a19799f309a297711d160b469eb59d7dd2d08f41f6aacaf761d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/codex-inline-judge.json", "bytes": 6057, "sha256": "6b5405dcc13a28154918a2a18786e6c6373f97a8e53effcde86d8f6de1089300", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-20T11-58-20-841Z-8fys/postmortem.md", "bytes": 9472, "sha256": "67de49c5999d1e024eafe9cf50a00a3777de8983a90ef97a0c228b4443cc8ed8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/539240c020.json", "bytes": 5717, "sha256": "a4c2c9d66e1c614a819c1e8bb97ee7aecf3ed25548dfbbe1cdafd4f7e913ae6b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "94e094e171033476d9560d54b651ac8fe1537d56e13d1695f283714925ca930a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2385, "sha256": "6d116625eaa062232907bf6812be00ded5009accb7f75eeb7de95d3acfd14134", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "94e094e171033476d9560d54b651ac8fe1537d56e13d1695f283714925ca930a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2385, "sha256": "6d116625eaa062232907bf6812be00ded5009accb7f75eeb7de95d3acfd14134", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/result.json", "bytes": 958, "sha256": "e5ecd02f4232868c40cfa3780586854fb2732b4690f7bc093cac20d99dba6473", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/score.json", "bytes": 3091, "sha256": "9eccfa5630725be07a04a1873489d9fb91620edc3211ea29cbca97fc5d12b345", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/metrics.json", "bytes": 1316, "sha256": "29f3d38f0e9026066a4533227ed57f46e6b0e189c264e0e1fbe8065e514b6536", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/facts.json", "bytes": 12782, "sha256": "3dba0a20601f414f6d5183696d3b744c9b4f445014f21ba6795e3e3dd4061505", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/codex-inline-judge.json", "bytes": 5868, "sha256": "3d03af68f0f53714beb7b79e4d7a6edbd6e38efc92c15363b71fed8a6b0f3509", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-20T12-54-47-060Z-srob/postmortem.md", "bytes": 7609, "sha256": "9fd64602ee338af5e637c6e2a820ae80de0b0dbe2325ef697243988fd21847d0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/6f8acf5e6c.json", "bytes": 4772, "sha256": "6048a787ed0b8285b33605ccf8f630b05cb5072d04b40efcbce21f787bb5138b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/out/customers.json", "bytes": 6231, "sha256": "d0e358d8ffc6b70ef3761c7d28372a389ac5338a4bf02303ef8d860613e90df7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/scripts/derive.mjs", "bytes": 2958, "sha256": "8e1563ee3f34b1e49fe4e5390c03b34a12465bd6b70092071e275d1e344c7853", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/tools/cleanup_script.mjs", "bytes": 4074, "sha256": "ffdf5b2a9312837a07fb37c4a61d35f411b22a244710046f1a4b22fd1bc86939", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/out/customers.json", "bytes": 6231, "sha256": "d0e358d8ffc6b70ef3761c7d28372a389ac5338a4bf02303ef8d860613e90df7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/scripts/derive.mjs", "bytes": 2958, "sha256": "8e1563ee3f34b1e49fe4e5390c03b34a12465bd6b70092071e275d1e344c7853", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/tools/cleanup_script.mjs", "bytes": 4074, "sha256": "ffdf5b2a9312837a07fb37c4a61d35f411b22a244710046f1a4b22fd1bc86939", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/result.json", "bytes": 1199, "sha256": "67f6c489d984314b910aabd6a7c4327a9ec5f3804b0b477dbb1df7f8ddb30c47", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/score.json", "bytes": 4243, "sha256": "be4ee448b4b7b1c7676fed5abc97fdcf2c5f1cdf2a39280167165ee55235abf2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/metrics.json", "bytes": 2391, "sha256": "69598346222489d5a3dfb52191053c7b07904ad1e35697d878d85f00b970acb0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/facts.json", "bytes": 13094, "sha256": "b7b1624c0ed736795d1baf2133d16fcda7059cfc0b2e5598b760ee3a3391c1ad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/codex-inline-judge.json", "bytes": 6373, "sha256": "9db8d01da07b1e4bd267e7e62e1e7d3cc7ad6cdfa5f52d2a08cec1862427cc1c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-20T13-55-11-836Z-ue4k/postmortem.md", "bytes": 9400, "sha256": "8767a0b2589b1e6b452b119951f3a6efe6284239ae36b1320da178ace55d0c93", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/776c87db00.json", "bytes": 7041, "sha256": "b984da2d253abd0defb91b02fd9cdd866299265c27a3ca8d951a625b2ad74dea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/out/customers.json", "bytes": 5110, "sha256": "94bd5a0ec084b7401646dd40bcd982216298aa602ea6a65c1270a931717d0718", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 5661, "sha256": "3193c46a76f86ff6644b90ccc6eb045cb42eba8d1efef48c85e270180ed26823", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/out/customers.json", "bytes": 5110, "sha256": "94bd5a0ec084b7401646dd40bcd982216298aa602ea6a65c1270a931717d0718", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 5661, "sha256": "3193c46a76f86ff6644b90ccc6eb045cb42eba8d1efef48c85e270180ed26823", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/result.json", "bytes": 1262, "sha256": "bfb0d1a01afb1ad7573bc29fedd9533cb6a9da61498c3235fecf1583f55d0515", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/score.json", "bytes": 4150, "sha256": "d622be7a8610454eac969867e65110c1d33e2a214a7b76821ad04b8a6039d9ac", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/metrics.json", "bytes": 2383, "sha256": "149bcd134cb5cfe044281ebcd621804fde77f3c722bcfb37ff77791e139b0ac5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/facts.json", "bytes": 12863, "sha256": "b9aa00a4931a893b1358a3da009d9e27365e99e81f9b724438b04dd45adf2bd3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/codex-inline-judge.json", "bytes": 6037, "sha256": "8a3d65b56afcdc08c41b0459890064debaaace32e54c12101f0966be25a54606", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-20T14-03-55-033Z-n1o4/postmortem.md", "bytes": 9239, "sha256": "b90cf963e9e091b292d7fac4d043cf67ef18dd9cfa71cae31f17e739633464a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/e6f37500aa.json", "bytes": 7923, "sha256": "6ba4bd2e40e0f599da3bde122679953db93a39747fff9ee10f1e7d98977aaaf9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/out/customers.json", "bytes": 7234, "sha256": "307ee509498c9b6d4503ac78c4baa22fc22daf1de5b0d150546ce915dc106e10", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/76e4077255-write-diagnostics.json", "bytes": 3145, "sha256": "ade9799aad571b32278fd07d6eca831f3c1524cdfcc0b854df4bbf8780b3b7fe", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/out/customers.json", "bytes": 7234, "sha256": "307ee509498c9b6d4503ac78c4baa22fc22daf1de5b0d150546ce915dc106e10", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/result.json", "bytes": 1287, "sha256": "c1671a2144b17e1c4bd7323e2b184dbdf31eed6db74f2e8d922a177b637be731", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/score.json", "bytes": 4161, "sha256": "6aab60fb20d70661a79756f5d6705c92060466ab99fef5816444a155850bc284", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/metrics.json", "bytes": 2329, "sha256": "a847e4acad12eaba31734430f29d870ff1cae6acc68d4148dbb396b8fa4facd1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/facts.json", "bytes": 12660, "sha256": "4c3a492c662e676914859d271c59c64de8ddc20a7eb07baa26ca14631c77fb45", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/codex-inline-judge.json", "bytes": 5723, "sha256": "219f6adad0142bee5949cf774cb999102a1d04a5d4676689b583c22c6d1b941a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-20T14-07-30-173Z-ez1q/postmortem.md", "bytes": 9204, "sha256": "de47f2fac55b9db2baaa33fb929e58563ba8edadea549c99c0af4c5dad0c620b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/76e4077255.json", "bytes": 5685, "sha256": "6325cd8e64c176de991631e18dc5245b767c667d421fdbe142101697cdd2a883", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/out/customers.json", "bytes": 6362, "sha256": "616a94a78796f8b224655fd68e6659500f695410970e012515558ddd3764083c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 2531, "sha256": "eebc555d0e7513e180eb09fa7eec285067c79bb49a32a61acf274497ada617dc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/tools/normalize_customers.mjs", "bytes": 3729, "sha256": "86f6f6c125852458c650f50d8d4106ed26f9c64b46abfabe80d8cdd4b1fca6e4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/out/customers.json", "bytes": 6362, "sha256": "616a94a78796f8b224655fd68e6659500f695410970e012515558ddd3764083c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 2531, "sha256": "eebc555d0e7513e180eb09fa7eec285067c79bb49a32a61acf274497ada617dc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/tools/normalize_customers.mjs", "bytes": 3729, "sha256": "86f6f6c125852458c650f50d8d4106ed26f9c64b46abfabe80d8cdd4b1fca6e4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/result.json", "bytes": 1512, "sha256": "58b89b58e460e232322f325fc2ca995506979dd773f3f4eea64eeaae7633d9ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/score.json", "bytes": 4548, "sha256": "96ec40c041383941f0008918167bbf709605e4af12453fff12014078e477d121", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/metrics.json", "bytes": 2391, "sha256": "5a081b3ee9f5044df34d083770556222b47ec17b647f69930f6853eecc34bd0d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/facts.json", "bytes": 19289, "sha256": "d30afeab497aad4b0ebd7ca8654b686288e587a87e2efae15b4cde561a16310d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/codex-inline-judge.json", "bytes": 6420, "sha256": "d707e1c526395eeb508ae22b6698db6ab14216d2c7b4ccf94d74aa2d2f1d7bd5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-20T15-09-41-929Z-aoqy/postmortem.md", "bytes": 9210, "sha256": "927dc2d243bbf52715d08dbee9588f1fb4141e9cf9673dc531817c8f67410a0a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/8996fa4f7a.json", "bytes": 6386, "sha256": "1781b6311cc8e349569a9be8b967b884ccafd14580f6dfdc3332603e704dc55a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/out/customers.json", "bytes": 5402, "sha256": "2d3cff0e46cecabe74078706838e560eba683647231d7bc03c63fac9a7040c24", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/tools/cleanup_customers.mjs", "bytes": 3336, "sha256": "b203b9bb2ef3667878aeec8cbefa0377bc398638544daf37ddc83f07155d9ef0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1984, "sha256": "5972fd4615b294ffae2f93886ede5b4bce7f6be9207553133180ea143eda2b23", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/out/customers.json", "bytes": 5402, "sha256": "2d3cff0e46cecabe74078706838e560eba683647231d7bc03c63fac9a7040c24", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/tools/cleanup_customers.mjs", "bytes": 3336, "sha256": "b203b9bb2ef3667878aeec8cbefa0377bc398638544daf37ddc83f07155d9ef0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1984, "sha256": "5972fd4615b294ffae2f93886ede5b4bce7f6be9207553133180ea143eda2b23", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/result.json", "bytes": 1287, "sha256": "2bf1481a3dcbf15435c2307d3028e8e0e8441efde3f9f0aee72709c637114e11", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/score.json", "bytes": 4160, "sha256": "4f279e4103aa71e1d5ba9bfede08069f216716621950965dc8b326196874dfc4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/metrics.json", "bytes": 2387, "sha256": "bedfa76bbb8ab48d5516ae563c99556b2a46b1355aff5220140357ed76bb0689", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/facts.json", "bytes": 14404, "sha256": "506c8c947e324b890fef88459685b1a6cc4012efb46497b323f549603f26d634", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/codex-inline-judge.json", "bytes": 5836, "sha256": "cc110a933d8ed42755751ec8a30e8ebeef39564912a8b1baa8fe905e1d2b7028", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/r1-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-20T16-11-32-055Z-7i1k/postmortem.md", "bytes": 9207, "sha256": "2503421dc4ad5250d4d0dbe9c83c72bad85f2024ea69345682958aeb85ad8364", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-2026-10-19/advisory-data-evidence/d7445ade6f.json", "bytes": 6902, "sha256": "529c61bbae36a0c478f450863ee5e7c8efa9bd775df7b5aa1038667c68ad3d74", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6265, "sha256": "01a8e2d38e0a24cadca5bfa4a7a404203ac19b62423e2a5f6d64810a70c5ea5a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6265, "sha256": "01a8e2d38e0a24cadca5bfa4a7a404203ac19b62423e2a5f6d64810a70c5ea5a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/result.json", "bytes": 1018, "sha256": "8d43fc03a4006a9511140153ffe3c752fe636a70065ce21d5fbea66258d618c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/score.json", "bytes": 3046, "sha256": "35dbb98dd4ec6decb25d4c54f4d2c18eef185f80113ee1e0c061984bdb79fcbd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/metrics.json", "bytes": 2388, "sha256": "f7e865deb7db7e242166a0f2bef943a2ce2b87db03f576be7233e4a02de44c8b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/facts.json", "bytes": 13340, "sha256": "8bc25d3ca9bd5ec4f6e19d46f1cc3862bf3bf71d8e49834c849522f38ab69179", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/codex-inline-judge.json", "bytes": 6324, "sha256": "d393c06636589a54ecde2ac48b6dca814dea044db236f7cb09338f34c648fba8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-20T18-24-36-520Z-b9xb/postmortem.md", "bytes": 7583, "sha256": "5ea27a721b850304aef48bec123a9d3137f44211b873ba0f23d211faa9864616", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/bf663c2b52.json", "bytes": 6060, "sha256": "89c2c67f0036fb73612a8e749bacb28864cbb6af45aa438dbea86d9164a01a2e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9087, "sha256": "7c25c788fb6b4ce29b50a34700fb73bd7321ea34ceba526ed8cec3bb09e512ae", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9087, "sha256": "7c25c788fb6b4ce29b50a34700fb73bd7321ea34ceba526ed8cec3bb09e512ae", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/result.json", "bytes": 1017, "sha256": "6d1ed134d712a86996db9e573df8e40f835003afc92d48c02e89e87105a1508d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/score.json", "bytes": 3071, "sha256": "01d48a9ec0583d443e62ad05a1e70f4578e2633ccc23c65045d48b76fec7e450", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/metrics.json", "bytes": 2383, "sha256": "be788cb62ed8e5eb9591b14c759f404abec284e5f4ca278fda17050b21616b8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/facts.json", "bytes": 15261, "sha256": "d758515fbbf320db20260d79bd8bb7b5c32599204a2fc8e06fe5bbb7eb92e3c3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/codex-inline-judge.json", "bytes": 6302, "sha256": "f303d326a31e97ef53aa5110a5236498445d089296f87a92d27f846b7e955727", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-20T18-31-27-317Z-mgu6/postmortem.md", "bytes": 7687, "sha256": "8c27aa53bf7a4ce4a80657c2ce3d6c34a00a1377c9e985f6c2ee0dc99883d28d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/f92e874642.json", "bytes": 6060, "sha256": "c6a27489a85be1057b4f99b1769f3fe89476115fccd81ba46b9aeae7ff2c3914", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 7755, "sha256": "759b4bb11e6a1f9d262d637788f10fd51d64afb3e25a5fdfb4451f6cbb3d9d7c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 7755, "sha256": "759b4bb11e6a1f9d262d637788f10fd51d64afb3e25a5fdfb4451f6cbb3d9d7c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/result.json", "bytes": 1017, "sha256": "6edc84dafaaea33c3c04587ff06466ac1bcfd4bea32c4e1aabfc51ab8f07aa0b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/score.json", "bytes": 3049, "sha256": "33a20b2d4bf05b725fd11ec45020b7ab696fe38635280b846ec575696159c194", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/metrics.json", "bytes": 1308, "sha256": "4e31f1cb94a0a6f9cfb814dc6895f71fdaa1e9804c32f708bd1314d1f976446c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/facts.json", "bytes": 11095, "sha256": "ae342a31e1b99d2c0f5a4fa1c67f885e6f36357ad89c74a7a004bccfc13b71a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/codex-inline-judge.json", "bytes": 6282, "sha256": "be494949be93305607a5e72991f6572859c1b5b1db7d48fb06ef4fe057ff6475", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-20T18-37-32-318Z-wooj/postmortem.md", "bytes": 7582, "sha256": "70b5b925b63c006eaacf8dacb07acb84cacdc287d0ff41fa10bccb8429266069", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/3ffbbc614f.json", "bytes": 6060, "sha256": "c9cb4dd98018a6dfc29831c511a81f0436fd8ed1b58032d20c29183517cf0393", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 10093, "sha256": "1b08da7a7cfd59c4ed51b0fac3aa5e2cc40aebf5f9d0e15924048b3c24465b64", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 10093, "sha256": "1b08da7a7cfd59c4ed51b0fac3aa5e2cc40aebf5f9d0e15924048b3c24465b64", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/result.json", "bytes": 1027, "sha256": "53389b39fe26073f30a7cfb2e8144e11f601dab4da4d1d2ca7e9034b5336031f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/score.json", "bytes": 3052, "sha256": "99710dd86fcb7f55b9c5f5a3fbc07324d43ff9e15537dfe12f36e9ed23465d5e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/metrics.json", "bytes": 2374, "sha256": "20e3b97af630aa6d01de5c91e1b23b33ab63776cccdc646ff56a01b8288428e6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/facts.json", "bytes": 13000, "sha256": "f9b5152ac4b978f1f0883fa4e1b80fc0492a011c03166dfbde42a53936137f1a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/codex-inline-judge.json", "bytes": 6362, "sha256": "c5d47b2b29f6b3680bec32a79714778810b34aaffabaaf9ebeb6111464a40928", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-20T18-41-34-271Z-vl34/postmortem.md", "bytes": 7583, "sha256": "bc774fe3dfa0433b8e0c2eafe23aca5c93d07e72a483bc36c181b769e35da40a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/379ff8202e.json", "bytes": 6116, "sha256": "bcac762cc83f59a8586e728f87fc3792bfc8af9acd8a173d1bd564d6b046a263", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9829, "sha256": "ff284d5df850cad2ff30a45cd849503e38bca4696455923524b28a0ff65bee39", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9829, "sha256": "ff284d5df850cad2ff30a45cd849503e38bca4696455923524b28a0ff65bee39", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/result.json", "bytes": 1316, "sha256": "c69666db084063a9da4d92d13b0d633e28429bcfb3dd5827b9dd4112160b3418", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/score.json", "bytes": 4461, "sha256": "a7aeeddf88f026d81e4d93a74ddc9634f88f4eaac9d10387a1d7fd3a63f8d904", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/metrics.json", "bytes": 2379, "sha256": "9df9b8f2d656a3b6135c138b4f17f4a48350d982d00a5c5993f151fc28c91501", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/facts.json", "bytes": 14591, "sha256": "cda37c24167c36783d7c90a50549bbb93a088c971a09d4996c4f2b3a9500efbb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/codex-inline-judge.json", "bytes": 6324, "sha256": "19a005dde291bdb0487804273edb52dd6be1da2fc5bdb6c93d6286670116875f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-20T18-50-22-996Z-uwq6/postmortem.md", "bytes": 9469, "sha256": "5ba89b905d65525c1505ba256df721f5943055e0df4e45fd261979e2c837fdae", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4ef8739aa8.json", "bytes": 6060, "sha256": "4858d08e7cd75d95b60b435205b8c008b1077e89e4f2818fc4ee30e5b6e5bdf9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9418, "sha256": "f26b1618426b92eb3703e7361791cfc570374faeea1583b3e782f58dd941b491", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9418, "sha256": "f26b1618426b92eb3703e7361791cfc570374faeea1583b3e782f58dd941b491", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/result.json", "bytes": 1088, "sha256": "40ee8026247a0c31d425ab17736e74e98fbb78f9b68a5a3b4d3781f78aac8e3d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/score.json", "bytes": 3077, "sha256": "e7b2b14b9aec574004f852e28bdaec829fce133a06a5cdd3edabc59a53515690", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/metrics.json", "bytes": 1314, "sha256": "edb550b6560c43a682db1e3054b53146153538c148b884c8a4260bfd6cb7f22d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/facts.json", "bytes": 10950, "sha256": "cd35b47cbb127636408688548de49bcbce4737fedc24abfa4a3f5bcb17d9daa3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/codex-inline-judge.json", "bytes": 6697, "sha256": "6e08305b1679019d4fe73db83d4d34466bd57a46329dd35d1f3f80e03fdb0b87", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-20T18-55-31-589Z-r81q/postmortem.md", "bytes": 7599, "sha256": "5cfe392e30b5722b8f778f57fe128dacaeaf0f4ec1c30aa374e20af988d70b9d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b6e4be8010.json", "bytes": 6438, "sha256": "c4caf1ab8c524c6a36dc2ad2167cd5497c7437548f829841d597b2ca23d7ff45", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 13815, "sha256": "d0a377d79cac664bf3acf9e3fd0ab17b5d5336473f0a3666b6c96ce7de338fea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 13815, "sha256": "d0a377d79cac664bf3acf9e3fd0ab17b5d5336473f0a3666b6c96ce7de338fea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/result.json", "bytes": 1024, "sha256": "4b059f1a9289cd9350358c5338f45be00bdf7aa9a13950f1b51185fc9d0eed61", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/score.json", "bytes": 3056, "sha256": "8ee6ec0a09d42c554d0c7d4c3a4b000ac9946bdff3b20faa58ab7c206351e828", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/metrics.json", "bytes": 1316, "sha256": "834410f2d79b227e1365b17b007ef2150da5338a229435510371333eb5e3b720", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/facts.json", "bytes": 11135, "sha256": "9b40ec1a74228efb100bc8afb29b8ea50852cbf18f27ff33f241889d34d57358", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/codex-inline-judge.json", "bytes": 5671, "sha256": "786216a6bc045fb1a1fc5c7d9274c2de05740a245c69dc0671d923f0f9ded47f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-20T19-09-34-031Z-r0ay/postmortem.md", "bytes": 7578, "sha256": "bb093442770bc18ac38495d78716807df1134e3d6c5adc7b9726dd7cffbd9764", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/877ba1515e.json", "bytes": 4866, "sha256": "4b25773b07686d4d6f74cc36b5c6b4386678d1a58adfc23a86ee9b904ab30814", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/runlog.md", "bytes": 364, "sha256": "d8e492fb292eba0c2a61fdfad2d2567943adf3f102e25df0963236b02b9d2050", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 269, "sha256": "4a54ad94f052bd36e3a32be830571bc09b5bd3a1d775e6cb6f00c13999138c42", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/runlog.md", "bytes": 364, "sha256": "d8e492fb292eba0c2a61fdfad2d2567943adf3f102e25df0963236b02b9d2050", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 269, "sha256": "4a54ad94f052bd36e3a32be830571bc09b5bd3a1d775e6cb6f00c13999138c42", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/result.json", "bytes": 1577, "sha256": "78d47ba83fd4cc5f17cc579b23c2a9378a633ba635103c8329782bb1675616a0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/score.json", "bytes": 4636, "sha256": "856399d08a3e3843026d03afa732b6ab1739f672e1bcd466008449bdbd91a136", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/metrics.json", "bytes": 2369, "sha256": "155daa8ceb9f152c6436612d54f231a757bcbe9558c87c833a45de8a75ce351e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/facts.json", "bytes": 15264, "sha256": "5d9a0f03a6c340944812ddc57fda6987ab66f12f6314d7f974937a22c8188275", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/codex-inline-judge.json", "bytes": 8040, "sha256": "4586eaa5c03c40e0d9e9e8183422d84ed9231a6de21559eda7abbe82ea4bea3d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-20T19-59-56-358Z-enbu/postmortem.md", "bytes": 9336, "sha256": "86f3bb422325d61e262c5dbabee18fead40772992d9a1bf6f30d154adb6525ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/235da7b09a.json", "bytes": 8598, "sha256": "9de44d38b86a3ca1eaa0c6a3443971ed5bc008c745a18a1fdfb894800bd67eee", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/runlog.md", "bytes": 594, "sha256": "b39ed6b7e5cdbb959c33e49ee7f50bf53d8e2a42b203d37261aad1ce45bccf4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 514, "sha256": "056c53dff4e4d7699046573f50973b5e34daa69df50d691e58bb8b5c3fa1661e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/runlog.md", "bytes": 594, "sha256": "b39ed6b7e5cdbb959c33e49ee7f50bf53d8e2a42b203d37261aad1ce45bccf4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 514, "sha256": "056c53dff4e4d7699046573f50973b5e34daa69df50d691e58bb8b5c3fa1661e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/result.json", "bytes": 994, "sha256": "39f77dad9fbaa80646dc6292e7ca4680f262ae05b921b43ff2b67da6bf29bda5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/score.json", "bytes": 2982, "sha256": "828d16d1d0cfa0f6a9b378f2447f08d7adafffc1cc2619046f82ef7e234e7c0b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/metrics.json", "bytes": 2325, "sha256": "32bda82ceebe2fa951800428a272e859ef15504f104d3cb29bcbb96100808713", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/facts.json", "bytes": 14672, "sha256": "83a5322b2e91d261fe8108a18b227e5c0bac5510666ae0c48d762cb99f34ebba", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/codex-inline-judge.json", "bytes": 8080, "sha256": "aa4a76a7181269702e0e72a1516f453a9624d56d3c27d7fc7ea0f3db63eacb58", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-20T20-02-36-445Z-fzsn/postmortem.md", "bytes": 7559, "sha256": "3afda2f235cfd28d9b6e6714d9333510b11a27b8e66447af6cf780e18e45dab0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/cdb69765a3.json", "bytes": 8482, "sha256": "3effffd2a493dd3a00d26c7cd28d73ca04b8e7b8f1135b4a3db9529ea33fb9a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/runlog.md", "bytes": 957, "sha256": "7959fe81d4f718d501cefe444eec124380551eb62e35c06169499f9190b19a79", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 883, "sha256": "05f211fba35193a3c12ab0ad53b4411b1f4587fd62e3920d4a74f336f814fa47", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/runlog.md", "bytes": 957, "sha256": "7959fe81d4f718d501cefe444eec124380551eb62e35c06169499f9190b19a79", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 883, "sha256": "05f211fba35193a3c12ab0ad53b4411b1f4587fd62e3920d4a74f336f814fa47", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/result.json", "bytes": 1335, "sha256": "4705f342c2718a2017f21229b1343ca7ad89e9e37bafa4bbc3b8c974d44ef574", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/score.json", "bytes": 4417, "sha256": "2c569bad3029d0dbae9dc9e0a9bedc2234c2102be98554aebd6f9f8d5410e7b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/metrics.json", "bytes": 2332, "sha256": "b7a51e6a7a038d20e30114304e8701f44b3b32c449f69e72b905ef3188ffa93b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/facts.json", "bytes": 16375, "sha256": "dc342d61402ce8c2efd96a3abef4b984734f810ac961e492ce10a8c63c072e26", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/codex-inline-judge.json", "bytes": 8366, "sha256": "6477b430536194d58e428bc8a21ce0dfe0dc7123320687e6bfcdaf96265d88d8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-20T20-05-25-754Z-elhj/postmortem.md", "bytes": 9951, "sha256": "33f7745b35da368f8a9391ab61b753c5fbd1f7e9ac20214309b8fb0c833985a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/77b92dce91.json", "bytes": 8523, "sha256": "f84daa4da8e5769993a69ad969f5d8b54b69bc7f4e61d6a42e4a6f2e1f297297", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/runlog.md", "bytes": 732, "sha256": "2a01a2f999657da6c1e9238827db7b5d561c66dce45bf90a8363ca9f349a8bda", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 424, "sha256": "7af01c1a876c2af54dc0e439b3c69534f9e846550efe239d79dbb03d79cfbc93", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/runlog.md", "bytes": 732, "sha256": "2a01a2f999657da6c1e9238827db7b5d561c66dce45bf90a8363ca9f349a8bda", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 424, "sha256": "7af01c1a876c2af54dc0e439b3c69534f9e846550efe239d79dbb03d79cfbc93", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/result.json", "bytes": 1427, "sha256": "3d1a413867df6364009e7d3d9f7e634531a0ba536b4cfe2f571e9e7fb8a3b2fb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/score.json", "bytes": 4390, "sha256": "018275340e8bc972e65b05facae043ae017f23d5962da3d02abb0fa3822f79ec", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/metrics.json", "bytes": 2336, "sha256": "8ca1f0047d5e77b29e6c347c37499c7fa2ad924f408fa557dd9fc064306c9a22", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/facts.json", "bytes": 14383, "sha256": "dd2b6e7d5c9f2667e0127b507c67357f3108156dfdb4e71730e749eaf84668a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/codex-inline-judge.json", "bytes": 8160, "sha256": "b323d835b96b28c1917fd86941eaf5862af5316714dd11dbad743eff636d3274", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-20T20-09-11-437Z-zdpn/postmortem.md", "bytes": 9278, "sha256": "c70adbc131ad1c4746924a17ca451b240f615af26932ea4b402b421c112a48c9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/57d232eb86.json", "bytes": 8598, "sha256": "eecf7344e0d02e6e69ecec00e3fc99f492f7ebd9c0b26a5588925c25d0d54215", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/runlog.md", "bytes": 137, "sha256": "715909b4cb3d0f96488dee6896f4025f9334b72455ec9d17b36b9850b6bdaa3d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 353, "sha256": "afa536a31e1fdc7efb7eb28bc41f0ae24785a83ae83ffc6763236b7654b16f77", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/runlog.md", "bytes": 137, "sha256": "715909b4cb3d0f96488dee6896f4025f9334b72455ec9d17b36b9850b6bdaa3d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 353, "sha256": "afa536a31e1fdc7efb7eb28bc41f0ae24785a83ae83ffc6763236b7654b16f77", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/result.json", "bytes": 1066, "sha256": "3258fef0a6a4ecf86cf1329dbfd1cc67e1584c5e1697b76857069bd16443d421", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/score.json", "bytes": 3010, "sha256": "f59dd0d67e02acdca4165f2e20c1bbfe8132af6e7a230252e26f79a072acdb89", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/metrics.json", "bytes": 2361, "sha256": "4c38a383798e0bf81ee9fe826482d0a6faa4ec37f5693a5d89db45b12d03d806", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/facts.json", "bytes": 14108, "sha256": "c755689f79d5de6353462895aa277932721408c126e84e502ee44fe25fd74464", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/codex-inline-judge.json", "bytes": 8553, "sha256": "4f025ce8d9d7a63d024239df56a4241cf2ed9cd86d2594c0cfb08d43c192b648", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-20T20-12-04-940Z-tokj/postmortem.md", "bytes": 7494, "sha256": "50d5772cb3aab030ef532d8ad59ff744e84a3837ddfcaba6735e84a7a499d791", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2b8e169b31.json", "bytes": 9084, "sha256": "6c81125471e4b067ca2adbb7de03524530654195ac544030c9accbcd45de8d9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/runlog.md", "bytes": 104, "sha256": "c71c0cdc428226643dba299d1b088b3ec2bdd5869b57d8a177f5c1c31a7e0cf0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 871, "sha256": "c3a861f392c68a070d01cd71b5aa1dc7f68265b50362dc52dd73187e45332c1a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/runlog.md", "bytes": 104, "sha256": "c71c0cdc428226643dba299d1b088b3ec2bdd5869b57d8a177f5c1c31a7e0cf0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 871, "sha256": "c3a861f392c68a070d01cd71b5aa1dc7f68265b50362dc52dd73187e45332c1a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/result.json", "bytes": 1001, "sha256": "8880a04a85051b7c7b9c7959fd7c2f8e8a7f07096586afafa1e4b2f1724bfe62", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/score.json", "bytes": 2986, "sha256": "a83334bac3005bb0b85d90876ca64fda63ff0179260d78df0c1032ccb8d7ce5e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/metrics.json", "bytes": 2328, "sha256": "bc571d80e6d8e5cae57673db6d5e4024e4e8530bab294fe285b0ef9605d7f41c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/facts.json", "bytes": 13999, "sha256": "eb68bb5669f89d61bc47ba1c9f4b6dc520ec50db404a7e4aaabb83e0e5644422", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/codex-inline-judge.json", "bytes": 8082, "sha256": "094519f7911d58408e2c5cb9e0c8074865c2cfe4ab04d6422d3de0ac6063bece", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-20T20-16-42-515Z-uqh7/postmortem.md", "bytes": 7530, "sha256": "97f78da707c8e8f7993592b2ff913410145831c5a0697547affe98dfac41993d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/ee03a8d6c7.json", "bytes": 8525, "sha256": "4379964be9a41ef3bc42f1af6d3506285bbbbb763351a402c796028d05a249a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/runlog.md", "bytes": 106, "sha256": "0b278d6548c92e8b36c87754cefce5f5bd8bc206cad23f5ff36cf5a2e3e249a2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 289, "sha256": "c1c22b5ea392ddcc364890e8788748d1a1587ef6e0219d8cf00eb128afc9b951", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/runlog.md", "bytes": 106, "sha256": "0b278d6548c92e8b36c87754cefce5f5bd8bc206cad23f5ff36cf5a2e3e249a2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 289, "sha256": "c1c22b5ea392ddcc364890e8788748d1a1587ef6e0219d8cf00eb128afc9b951", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/result.json", "bytes": 996, "sha256": "4fff4e1f743e2fba14d7f0cb2bd05e1b2b028f633674603c42b8ccce3fdbe5fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/score.json", "bytes": 2982, "sha256": "76a76e996a091ddb0c40007cd6446779005c41301269ce329c774b5bef3583f9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/metrics.json", "bytes": 2372, "sha256": "7bb79f0227b1fed8fe2d914911cb21d490484c09221743f8b88fa172ba17f781", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/facts.json", "bytes": 14090, "sha256": "a8265247249f478b38d29bf454f82ab4c9a69383934ffa56c0b0c28be26b1ea2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/codex-inline-judge.json", "bytes": 7990, "sha256": "7d9ce595150d197db860e30551dcf78f69ea03c4453b28bf07835e1260385b51", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-20T20-24-44-705Z-m6wi/postmortem.md", "bytes": 7476, "sha256": "519c1bfca790d73de73648fd5051868413fdf61071e725e510d4be2fadc13f17", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/32c509ac38.json", "bytes": 8482, "sha256": "4be75d8bcce0e1801780cd6f43afd0a322416a2885ac73ae10023a927d5b57dd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/workspace/harbourview-office-relocation/plan.md", "bytes": 3529, "sha256": "add6b07b194c38b9f88f0bd3ef6bf30291fc49ae52ffcfad49ddb4c94d6aee6f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/workspace/harbourview-office-relocation/plan.md", "bytes": 3529, "sha256": "add6b07b194c38b9f88f0bd3ef6bf30291fc49ae52ffcfad49ddb4c94d6aee6f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/result.json", "bytes": 903, "sha256": "7974b369aa90679c149974df402b35fb8216587237e65f97a0b04a228cad9697", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/score.json", "bytes": 2903, "sha256": "6ce5600a1b9af11fa87a344d7e6235a93947ab8f35227a9c6a0b2f97a0de823a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/metrics.json", "bytes": 1303, "sha256": "46defad4549daac19b4ddbc2f0732c480b688d5b55d17e055b2d7d9a828a2356", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/facts.json", "bytes": 10093, "sha256": "5a339fad92761352e58e8b043d6ef08550411c8d452f5a9492c10a00f123dd92", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/codex-inline-judge.json", "bytes": 3988, "sha256": "e42a61268937117b423e34f401ab1f20cc4c3d548a784f1d11727e75d3af5bb1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-20T20-27-47-933Z-g12l/postmortem.md", "bytes": 7526, "sha256": "12a40c9a7f6eaa54e505523b8084f7ecd706f994f3cdd5800b6394142d586a0f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7a48e4d339.json", "bytes": 3658, "sha256": "5510361b8bb7de90c272678eda2a49481a776116547e35ee52b4eedc64d92b31", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/workspace/harbourview-office-relocation/plan.md", "bytes": 3301, "sha256": "701418fbd145ea827d75ea022e089df43ba267d749010a5f1bdbbb94cfabce72", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/workspace/harbourview-office-relocation/plan.md", "bytes": 3301, "sha256": "701418fbd145ea827d75ea022e089df43ba267d749010a5f1bdbbb94cfabce72", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/result.json", "bytes": 913, "sha256": "05698eb21af4d55fdaca38744d119c4ee6f09df8499f5be10df8cd64a7f8edc5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/score.json", "bytes": 2906, "sha256": "8773d81eb32b7a5a33dd10e9c15a4621db3e7acff30f2d5bc6891ee5e83e4a88", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/metrics.json", "bytes": 1304, "sha256": "c219084a817972ee3e70244709ee92880da10f2d89db8280decd2cafb929be94", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/facts.json", "bytes": 10133, "sha256": "1c4452227f99b943eaf8f8f85f2288c16c5bc60fac66226aa204a940bc8a06f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/codex-inline-judge.json", "bytes": 4024, "sha256": "156dd838c8d0f7b40cd0ccc4a5a0b47481cfaa3ff3fa12ca84a3378a685e0b83", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-20T20-29-48-642Z-zv1w/postmortem.md", "bytes": 7530, "sha256": "6270946ab805299f910372ce122aedfebf8ff0c6e72613ac1d3eb3a59b3af924", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/3d6f1c04b4.json", "bytes": 3692, "sha256": "b6bb6e07a9093b4342c2aa691ab3c102701f911edee7cbec374f7433df772a9f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/workspace/harbourview-office-relocation/plan.md", "bytes": 4695, "sha256": "54bad8053a5d8d6625b21f1bf72ef9e6781d02bfa9a53429c6bb3f29283e5035", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/workspace/harbourview-office-relocation/plan.md", "bytes": 4695, "sha256": "54bad8053a5d8d6625b21f1bf72ef9e6781d02bfa9a53429c6bb3f29283e5035", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/result.json", "bytes": 902, "sha256": "61ba1a2fff7525225d86d28f81d1bbeffebaa3b11495206acb2ffc7ab4a7c459", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/score.json", "bytes": 2890, "sha256": "c2743288fd148f4a1adb5f980da13d1eb92f99713b1ac2a33a3cd56c42e1c6ad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/metrics.json", "bytes": 2364, "sha256": "df2d9fc0fe3a8ab2f9b12e36ba18696bb05d75a07fc91a03cd44599de88c8e26", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/facts.json", "bytes": 11286, "sha256": "293edcb4e1d6615755672ccc0cc6c1be551cd1a991990a603cc3e75f5fba7ae0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/codex-inline-judge.json", "bytes": 4006, "sha256": "8cca0ad1c509818829cc5137e41f8cef704e8b31a3716ccdefb00b4902a02bc7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-20T20-32-02-153Z-bnhb/postmortem.md", "bytes": 7506, "sha256": "14dee22a40faec8a895bccd33e6027edd1d6effe0913dd2ac230551a7c9d9a9e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2fd3230e10.json", "bytes": 3658, "sha256": "611bb3be10e67bd57e03fa7e2fba1ffb8684e394a2af7adb7aebe25038a23b5b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/workspace/harbourview-office-relocation/plan.md", "bytes": 2941, "sha256": "605d9451d10ecf53893c9650d5c3c0ceb2ce90ffd6e580bf7acb76883a368588", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/workspace/harbourview-office-relocation/plan.md", "bytes": 2941, "sha256": "605d9451d10ecf53893c9650d5c3c0ceb2ce90ffd6e580bf7acb76883a368588", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/result.json", "bytes": 975, "sha256": "507f6c6c6913fbd5f9eee05fcd2081d65332105360b9bd53d109dc97c1b47b8b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/score.json", "bytes": 2933, "sha256": "2c2625d163350a6307f316a107750dacd71d9275d9ae6c97e3d899e65588603f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/metrics.json", "bytes": 1311, "sha256": "60382f61a1f43274a49c7f7c085cde29728fc44e556b8509467dda4f9e2986b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/facts.json", "bytes": 10358, "sha256": "32a954823684d290a5227b6f1edc700541774e5c9108702234199d748bd5837b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/codex-inline-judge.json", "bytes": 4225, "sha256": "600280f2de3616856621e4c3095ce3169c751cf2257a2c3f6d7ca4cc000c76f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-20T20-33-56-623Z-upar/postmortem.md", "bytes": 7564, "sha256": "f4cfad4629271e379d27812ab12d565d2706dfb8b65b99f39f2b1dfa7b4de072", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/101e942882.json", "bytes": 3896, "sha256": "5db2809ec5a69e0d4e41dd3c314ed968c41e8d2b7155673c3db896f508b99f7e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/workspace/harbourview-office-relocation/plan.md", "bytes": 4912, "sha256": "db0f43590c387a50e20bed0b49370310740c3430f252517cc1ad966195d3a56d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/workspace/harbourview-office-relocation/plan.md", "bytes": 4912, "sha256": "db0f43590c387a50e20bed0b49370310740c3430f252517cc1ad966195d3a56d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/result.json", "bytes": 910, "sha256": "874a66383062a02b08819c56d82300d3d4d2e52db789c0d3fb0ecb9203d0b817", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/score.json", "bytes": 2907, "sha256": "703eb8070cd73b5c32604a010fde668f060f67e8333ecdcd3f829e502c937109", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/metrics.json", "bytes": 1309, "sha256": "950f9a8054d398925ae29170cda9f1568df4df6ea9401e2c35955df819008e9c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/facts.json", "bytes": 10251, "sha256": "785ca1375bc5e58a2c436d5e4ea2615cf48372c48ccbe6ed235336d177246d57", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/codex-inline-judge.json", "bytes": 4015, "sha256": "bbca5fbdddacf884f5a4f21f86e8b5cf8a530a883e38b379322800af9aee450a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-20T20-37-26-847Z-excv/postmortem.md", "bytes": 7545, "sha256": "cbece7443415b9bc75db416c82b6cda4255b54dec7a65d8f2f8eedd0c78f28be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2d64808227.json", "bytes": 3675, "sha256": "738718471cc7825a255c17fe1af21b186550f9222ebb1a8a3786da6620c58b3f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/workspace/harbourview-office-relocation/plan.md", "bytes": 2418, "sha256": "1a2a00a9b38262215c8ca2e6bd51fb657d3af4c90df8e35cdf26963f6d1e0976", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/workspace/harbourview-office-relocation/plan.md", "bytes": 2418, "sha256": "1a2a00a9b38262215c8ca2e6bd51fb657d3af4c90df8e35cdf26963f6d1e0976", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/result.json", "bytes": 904, "sha256": "b57edd7edc054feff44483f17134b74511bd807f04ad95990f3df1148563cd4c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/score.json", "bytes": 2903, "sha256": "a4239927db57d64169becf2abaa8325777fe5b958282bc34619e1d044dc6fb67", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/metrics.json", "bytes": 2371, "sha256": "b79daf7cc13e777d794d41936688448754ff26dca1c381adbfc2f479b5b60602", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/facts.json", "bytes": 11354, "sha256": "fd1014c0b11a620146a790a1d31ef0c0905e77f1235dd5f259dfa389bb997749", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/codex-inline-judge.json", "bytes": 3964, "sha256": "a6f2f01b9324d8f2ad1fd79d12b2e5d2cd3d96cdb49f1f63660c9e0174e604b4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-20T20-45-43-125Z-tivu/postmortem.md", "bytes": 7505, "sha256": "ba8ed46cddd8e74c2a29c38abf2eab186ca57e502d0af73d318522e873446cd6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d5dcb23be8.json", "bytes": 3658, "sha256": "14c2a94b3b59a1f2f28613b3d33a3113445fa99292a4f33c1cba9a1326365668", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/workspace/harbourview-office-relocation/plan.md", "bytes": 3131, "sha256": "ac148c68473956118d31aff812116f8ae7fef4222371c76a554e84216654ee99", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/workspace/harbourview-office-relocation/plan.md", "bytes": 3131, "sha256": "ac148c68473956118d31aff812116f8ae7fef4222371c76a554e84216654ee99", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/result.json", "bytes": 1224, "sha256": "48abdb069810be51d08a9a9ee7200c3da28e1f8e0df3ae9a7e04658a4e92d11c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/score.json", "bytes": 4095, "sha256": "c5adbbf08eec94ccac616236fd378bd0ab0d7b9c31612493dda9c2ff9aa7552a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/metrics.json", "bytes": 2329, "sha256": "3fcfa712193d8844a22394ad32d17ade270ba1e8a04fcab8e85e6e5721302f03", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/facts.json", "bytes": 12544, "sha256": "9dfe26ba153aeda2eb3151b652891e95fb13da78f2e662cd78a6826e63510b0a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/codex-inline-judge.json", "bytes": 4056, "sha256": "fd307acc4230432a98b759753d4c7d1ea991c8b0ceb792143a498c873b201265", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-20T20-47-24-043Z-e4h7/postmortem.md", "bytes": 9922, "sha256": "260864487e9eed368dffbe894b1e9ddda0e8e62801bef2630a96b5b1dcf4174a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/06a1853983.json", "bytes": 3892, "sha256": "90eafa793cdc24c98c8768da20230b5f4eb450eb4f16fff89c19dfe4b7b4feb7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3565, "sha256": "00d5f4855297ea2de672dd682dffcb1f54f75087d39a76b3410bb25e55a51f15", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3565, "sha256": "00d5f4855297ea2de672dd682dffcb1f54f75087d39a76b3410bb25e55a51f15", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/result.json", "bytes": 1718, "sha256": "eb432474e16b0528cf58baafaabde471f1143a6cccaac453253935f5077e71e7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/score.json", "bytes": 5507, "sha256": "9823bc10684dc86cf19864b98bfebd763684f69f38b82b5406cfbd8e64f85b52", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/metrics.json", "bytes": 2381, "sha256": "dc486939cb89a7b19e3ca78da720b51f9df74f64b2c3e5e1a8ddfe4da3e71550", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/facts.json", "bytes": 17728, "sha256": "6ebe168463ef05bbd7f2d5a765fb012e97254543588bc6b7cad977df742b8464", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/codex-inline-judge.json", "bytes": 6634, "sha256": "31d51879794443adbe10f4e1183854706f13b8597a7aedc961915521fc8162da", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-20T20-49-46-562Z-m5y4/postmortem.md", "bytes": 9376, "sha256": "b237df2a6efa70efb20498d6db77bd4b627828f7cab4018192d8bf117f26ea75", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/422b1d0552.json", "bytes": 6020, "sha256": "9eaef40da0a64c00f36a025fea4aa13b1cc76e1a565f89d6baf1951dcf523c1d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1850, "sha256": "8d9f353eabf54ede2459e324ff305d8aef3f053f5f74f757f79dad327c35f24a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1850, "sha256": "8d9f353eabf54ede2459e324ff305d8aef3f053f5f74f757f79dad327c35f24a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/result.json", "bytes": 1626, "sha256": "80166f28183114829080559be00d997c0640917438d5c645aa8d1454d3f22bfc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/score.json", "bytes": 5479, "sha256": "766f143c06d542fd84979ea2874a68b98ff4aba538c4f5a90bdeb919cefdeee6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/metrics.json", "bytes": 2378, "sha256": "7128496c46ab3eaf590c4f5c3e65a98d40a011264a7eef455e8a0e8b4482f5da", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/facts.json", "bytes": 16376, "sha256": "2236d5baff3673782151021494eea9bf5b7998938155ef3f9132f13742c91e09", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/codex-inline-judge.json", "bytes": 6576, "sha256": "293eaa3c737d45df29c3c5efd643b33f6a1d05eedd164f0a51ff6bf4e541f22d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-20T20-56-24-260Z-09ai/postmortem.md", "bytes": 9366, "sha256": "7785c33191ee272a76932dd14a698851c5c753349dbb0f7d64b7c6303b06477d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/e092513fee.json", "bytes": 5954, "sha256": "a876a78472e9a4489a4d6ee49bc7c55ac1b440146701627436cac690642f5b37", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3955, "sha256": "e8b94145a4211672143713a0f1c4023814185dc18243e9047e08f943c35c2356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3955, "sha256": "e8b94145a4211672143713a0f1c4023814185dc18243e9047e08f943c35c2356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/result.json", "bytes": 1247, "sha256": "abee6482aebcc1b5b3776e367b085f5b171e5810967c2e0811715b0e3d657a05", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/score.json", "bytes": 3332, "sha256": "1d277756bc329ff690246a8f74fb02293a901536a9b9771495fbe57182471999", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/metrics.json", "bytes": 2412, "sha256": "fc0749b38712a01f352ddec961903a352439dad3ecc35c91396022711bfdd95c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/facts.json", "bytes": 15950, "sha256": "9a46ac1695012c2deea10edbd9f741338c42d1b207dc648a34d0ba0ebe4abcd6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/codex-inline-judge.json", "bytes": 6971, "sha256": "8f128a2a7cb1b7a18f977ffe1e0cb5e1b2a550e1499765346b1f856983876f28", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-20T20-59-19-544Z-l7hm/postmortem.md", "bytes": 7694, "sha256": "89f96af13ffe610f1215a816b03d6cc45cae1489a73f46fbbbce0f86ec884793", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d2f7712840.json", "bytes": 6416, "sha256": "dc17a614bfbe6ab9a62e1c53286066371fec54fc88c9940dfe6efda42ac58c67", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3105, "sha256": "738eb48618c219bcb47c19fce7d25ad15a796664747e1c72edbbb87ea587ee09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3105, "sha256": "738eb48618c219bcb47c19fce7d25ad15a796664747e1c72edbbb87ea587ee09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/result.json", "bytes": 1182, "sha256": "25b5ec356d2b6c909c84b447c06466183dbfc0a5899dfaa89884fd2d0cc65815", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/score.json", "bytes": 3291, "sha256": "9093b1b53e4f6efe053a5d8ad9e01d68d31b0c70d0be211e2534444dca8e73e7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/metrics.json", "bytes": 2375, "sha256": "42213d518b9c5e8f8e62a209c70c2897060994f823a0f59ad35a40b3247ed8f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/facts.json", "bytes": 14892, "sha256": "8f606f7a8bc58128910259e898607047cc0ec5493c8161f753da51c7ae547704", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/codex-inline-judge.json", "bytes": 6546, "sha256": "e97ed320d02b9b3f646e0f87dd00728bd9d5d758b5288323d31b9b9cf381c0f5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-20T21-23-33-475Z-h7dl/postmortem.md", "bytes": 7591, "sha256": "067158c539fa740f0b0a9564b1d27692adf6639c83ad67386fb9586cc80e04c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b273bb308f.json", "bytes": 5987, "sha256": "c2e5fd0b77ebd194fc4a1b52468d3212eb0609098bf466ce07c2b95e6ff1aaac", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1537, "sha256": "dec6b750e14df537fba6e90318d2588a157e63283a72f3ae50791736b756a8e2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1537, "sha256": "dec6b750e14df537fba6e90318d2588a157e63283a72f3ae50791736b756a8e2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/result.json", "bytes": 1175, "sha256": "6ce21dbcfc718dda4ab9d6dc2727fb027fc98d9144a86dc0cffd83754b22d754", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/score.json", "bytes": 3278, "sha256": "1593fb01bc64a856e27b7a29937b9d9f9ec35c42352acd35824a0c888b139283", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/metrics.json", "bytes": 2368, "sha256": "4a47d599494118e7c11f55d8d37a0844256e3ee32f0d6e892489517a94089d84", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/facts.json", "bytes": 12855, "sha256": "7f1625e308e1e1d8d0ab37da3484903077226523c647aaf4e9f68019d1a19a42", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/codex-inline-judge.json", "bytes": 6517, "sha256": "6d7f84d6f44151e4a22eb26dd9bc7b220641f455f1b16f89bdff3aa2c185e769", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-20T21-45-20-753Z-5i1g/postmortem.md", "bytes": 8023, "sha256": "81aede41792ba9c7e5e7a7eaccdc3fa4ec5f9daaf84cca7955a072b216d15619", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/96d61bd46c.json", "bytes": 5954, "sha256": "a94271e6622695e543938b0aceb4d373f4dd0e3034a9651715629ca593379da7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2154, "sha256": "aed9817c3e0d6ce50e6d506b6631c6e1b9a12a07bbc5928e0140fe563b00bfea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2154, "sha256": "aed9817c3e0d6ce50e6d506b6631c6e1b9a12a07bbc5928e0140fe563b00bfea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/result.json", "bytes": 1175, "sha256": "1b57315e14be47839e8d391fae7ce80e00ed6a16d6b08a08ad1bf0793004dbcc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/score.json", "bytes": 4528, "sha256": "2da0ed6f13b2278aaa218db13d148722a6ba2fc7124a0cec25db8d13bdd1a533", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/metrics.json", "bytes": 2371, "sha256": "fc0785d7c2858bc2577849f5505ef33e5caaa73d31b3c13a8c774f21c3a12b67", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/facts.json", "bytes": 15850, "sha256": "ca522ed1be8398269ff3485c8fcbb0a0e956fe07199dd0a0c3e1cb98b1376f92", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/codex-inline-judge.json", "bytes": 6505, "sha256": "0c2c6d2be1ff85056455942e0366f41ca10a9df342b07d63b4c21417e45bb112", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-20T21-47-52-307Z-xcdy/postmortem.md", "bytes": 8085, "sha256": "1f287fba45955ec0e328247c1c5717a3f6f37fcfa367d9c3408bbbde5085a7e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/dd5866a97d.json", "bytes": 5954, "sha256": "efefcd0ba6b9f3743d295b25469180afedc403b9e34e5a31ada872ce6506fc6d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2448, "sha256": "7e7bd7d51c565e17cf9164a807a3d47c745f87f8325bdff864da644df4c0582b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2448, "sha256": "7e7bd7d51c565e17cf9164a807a3d47c745f87f8325bdff864da644df4c0582b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/result.json", "bytes": 1174, "sha256": "84c5d3765e9cda1260b4a0b2e355ce00cfb2960790b49e452a4ab0a1fbbe5657", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/score.json", "bytes": 3265, "sha256": "7213331f587837aea3f74bbae1cc515a005e8820ceac2a009edc65d364ea75a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/metrics.json", "bytes": 2367, "sha256": "503a1fd440ac99dd5e299364564a18aa72cf1f458c863e88351afa6662977b3b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/facts.json", "bytes": 13870, "sha256": "5e8e4da8139ddad2017b7cf6f66938273ec8fd3ba68cd71dc4fbeba46adef70f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/codex-inline-judge.json", "bytes": 6483, "sha256": "f994e11ddf87f39b06919fb17d8bbce874f8d2663761a66271b2ca0bd90ef363", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r1-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-20T21-50-55-681Z-ta4y/postmortem.md", "bytes": 7492, "sha256": "2f8ecd1d713de5e0672ed506295e6cc77b1ce37260afbe841d4cfa0f5cf12e9f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6d35f8b382.json", "bytes": 5954, "sha256": "011f4dd84c482360443fd14ab60034d1c4eb0ef1c0c07f10b5da15f0973e56db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/workspace/tic-tac-toe-game/index.html", "bytes": 3728, "sha256": "6253f8a1cdd1c0c58b252b909fe25e3f367f910d9e6657faaf8312b65548bcff", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/result.json", "bytes": 860, "sha256": "7c01628f80447eccf891b2e6e5245f7f14859ad34922c7c183792c6109828a0e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/score.json", "bytes": 3297, "sha256": "aff12b600937fd63e09da5615c24795836247309e6411e5dbdafeefdf50b9b9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/metrics.json", "bytes": 2372, "sha256": "dbdf6e3e78f29bf8e6cb54fed1116c562131b4b3766d1ed5f57d28679df95f8d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/facts.json", "bytes": 12085, "sha256": "c646615859029d765e09322ca2265cb56e88b498727d28a65cdb95b51b32e52f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/codex-inline-judge.json", "bytes": 3305, "sha256": "cb0a8e7bed3a0b2bf985f1a3d13763084b93a1d7dba2c6124c60ae20d20c55d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-20T21-53-25-605Z-o447/postmortem.md", "bytes": 7861, "sha256": "faf62c6bc1842720a4037c75cb979671ce08fb53a4936b51940feb1aabfbe500", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/109deb2e74.json", "bytes": 3910, "sha256": "0c11dae0bda72f1b4519126590b04a7fa06b2b63d9d13b0131282e3c4658bd91", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/109deb2e74-desktop.png", "bytes": 10584, "sha256": "a54b33cd79ff74fa118f2320984d24e96c59d4202eceeb9131684ae74980a0bc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/109deb2e74-winner.png", "bytes": 13033, "sha256": "85191be6d3043a6b9310fe92520e175ff511a5f35bc2febb31bf3006bbef421e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/109deb2e74-mobile.png", "bytes": 8149, "sha256": "b1a7ac414bbd899ef65ca41390aee35e08ada2bafbb691cd1076f5ee4cc7b031", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/workspace/build-tic-tac-toe-game/index.html", "bytes": 2339, "sha256": "58fbb49c22f2f8bd8c69b94415bb7258b499edf53bd2820c6528cb401a20aaae", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/result.json", "bytes": 867, "sha256": "940bcf7908b43cbf767ecf60fed40dda7c2f5cfd5b02b863d6d6c92acf58e65e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/score.json", "bytes": 3321, "sha256": "3e0914be008335004c51c7a59cb18d11ab11c3bf2d3e77c8b16553a2720123dd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/metrics.json", "bytes": 2377, "sha256": "20b0e9099fa1d54ae8d2b7873562d55702a0f455b158e9517d39530344a8f5e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/facts.json", "bytes": 14311, "sha256": "0941ca1ddbaf27af2df3d2752af4c0e7ad53915c61ea30d15c2e7b4b023d1851", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/codex-inline-judge.json", "bytes": 3291, "sha256": "e98ab19a7e76c9a74086f376a4570de96730c54baa3c87fa69a3dc876d54fda2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-20T21-55-56-777Z-wznr/postmortem.md", "bytes": 7907, "sha256": "18357ad6f2a9b10a25831f6ff1709911169274b58a2a946fa4744df7ed7698fc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d1ac09efb4.json", "bytes": 3915, "sha256": "d74d9461eb6436b8f7c0745bc9c11152e532654e72c13e0219423c8f3ef0da9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d1ac09efb4-desktop.png", "bytes": 8544, "sha256": "1af0a42a6efb00bd0d11112a95325a530d4cbd98ad5a77ce0b6ec85a43b11628", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d1ac09efb4-winner.png", "bytes": 11920, "sha256": "fae37f4df0e9e687528c69ce5c9541f4a9d6f201759d742c433ee6c08ffdc859", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d1ac09efb4-mobile.png", "bytes": 6158, "sha256": "36ddc154279b42e04ea2fd0855dec222ccc6c780847f94f73f1d5e4150a113ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/workspace/tic-tac-toe-game/index.html", "bytes": 4900, "sha256": "e0a4e8ceebad7ccae877fb74ddd82347bf06e72ede0a3cb94b177245e2075247", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/result.json", "bytes": 871, "sha256": "93dce3138e0bb6fdea50c284efec5dc23577ba83ca6ca8366b9a35cc3cecdd3c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/score.json", "bytes": 3095, "sha256": "1c1b7f2bc0d311e2a926069acac9c95aba1f54a3818dc6f97330d9bd678dc5bb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/metrics.json", "bytes": 2377, "sha256": "9b3c55506b86184f51fb7de426a5bdb7753b8a52912af41202f50537ae4e2611", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/facts.json", "bytes": 12520, "sha256": "34f117884170f711d00ca8889d00933f5f1a8434629f70563d21e51406a1bd4b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/codex-inline-judge.json", "bytes": 3326, "sha256": "846a24edaac001fb09e92fd4e0bd643a45b6dca2836aeb2c516dd7687df8b6d5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-20T22-01-31-804Z-71i2/postmortem.md", "bytes": 7667, "sha256": "42b835d3e1cf65e81f447273870a7786b21dafe61e1424a69df72023f97909bf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/616e16a003.json", "bytes": 3973, "sha256": "06bcbac561329f10abb37961411297f21f87855c9beeef74edae29b00b3431c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/616e16a003-desktop.png", "bytes": 12283, "sha256": "f205a5d155a13f5610e5cb98ab10f8177e5136e83c74d4e9326dee38449a7f03", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/616e16a003-winner.png", "bytes": 14933, "sha256": "e2aba31bb6abfd1ed502c62afdf25d60a2f96fd391873516fc4116d80de23946", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/616e16a003-mobile.png", "bytes": 10062, "sha256": "50289824b4868ada7cf828baedfa4160364965d8886a3062f4e766702324a0da", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/workspace/tic-tac-toe-game/index.html", "bytes": 4713, "sha256": "aa7f78cb0eea2de4b3a337e7f9e2870f7e8c1c263697f7731ec3e23369b13e47", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/result.json", "bytes": 859, "sha256": "ea67c689fd0a29e1b2c916cee7ad6fa798bb3f80aac6fd692c71953a167e2c6c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/score.json", "bytes": 3088, "sha256": "d044ce30bee9d37834bb343a983a836a74f883ffd5e4a6b0de555d3dac14fe2a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/metrics.json", "bytes": 2334, "sha256": "b5fb686262ceff2679c98b5776f51e0b079e0449c65421a1b8bb26b7b7ed01cc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/facts.json", "bytes": 11898, "sha256": "3b528ea89892d729b92f7cc7834c21d9fddca8a9c8513d1383f2a269e2c94346", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/codex-inline-judge.json", "bytes": 3277, "sha256": "2f3b19f88595855991a4b1c3d1a368f4925d88ac01cfb8be15602fee1682cb08", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-20T22-05-52-660Z-51hr/postmortem.md", "bytes": 7621, "sha256": "d2beb5b3323d4069b787660e648c6dfb6eede5cc30b5dcad228d39b642e8f458", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d197c56107.json", "bytes": 3924, "sha256": "382c4bd742571acbcc4d6d72a67374c41917c199dba9e269bba1434b4e08b5fd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d197c56107-desktop.png", "bytes": 11516, "sha256": "6bc9adf0e7f1ea19c9b17f77951f7014ef50ccffefed537a7d63f2adda10df1b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d197c56107-winner.png", "bytes": 15384, "sha256": "71e2079710c3990e53d57848388879f06d37654cbd2ca66bb3bb3c9bf029833e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d197c56107-mobile.png", "bytes": 9033, "sha256": "6f6bbf94afc0a1667182d04066682172aea97c8869d504cce41ee405b38e2509", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/workspace/tic-tac-toe-game/index.html", "bytes": 4131, "sha256": "fdbf91400bc77f548fc123e7c93e9235180afeaa10d9148a844b36a0bf238c4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/result.json", "bytes": 932, "sha256": "247bd2da48e234dff6b80a4d9ccb3d3965843d756444ae7eb796c886f0cd5ae7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/score.json", "bytes": 3118, "sha256": "2ea02e53dfd62d25952c2ce897df4c90c3abc6602d9fd1c26f3206ffd24eda2c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/metrics.json", "bytes": 2403, "sha256": "699c195a33b601677002c60eed28516f6545c3f903c7478f02403c85d0b338ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/facts.json", "bytes": 12078, "sha256": "95496590b8bfcf9b91388a0cc110c5360726d960f9a1c0ec30ea46ca2afce6d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/codex-inline-judge.json", "bytes": 3130, "sha256": "60810b0a5e631e036f12fb3c7f2b62792312b30ebb46137d68b0a6f0d48eec54", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-20T22-08-23-181Z-r72g/postmortem.md", "bytes": 7671, "sha256": "7fe477409cc60bdb7fa8200e95ab47e472deb5f7061d2b309b2de23405f97c0f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/89e452db6a.json", "bytes": 4123, "sha256": "22659a9bc6ee00715db2a263f893cc206371caac0e1806e1eaefa6399e6d1d0a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/89e452db6a-desktop.png", "bytes": 13966, "sha256": "fd97ae5c7338fd2417c96dea10d8a2a15f62b03c66b66a9237c0fcec9c751a52", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/89e452db6a-winner.png", "bytes": 15556, "sha256": "fa3fb717c583c609d3717ed5fa6e09937aa0bc522cb5fd2a49fa55b5ac753b6c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/89e452db6a-mobile.png", "bytes": 10933, "sha256": "a9730a052bedfa80c9371370447cbb9f41ef29c26f6ef145a9dcb8c91024477d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/workspace/tic-tac-toe-game/index.html", "bytes": 3356, "sha256": "9ab0d8192c061c0253821b91c23013630aabd911dee34e10b260f55b3e5e4ffd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/result.json", "bytes": 867, "sha256": "068eaf95d2848850e87731814ef2587ca26572be91b325596072477cf4852774", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/score.json", "bytes": 3302, "sha256": "4e07a2079e4a631eaf961ef96413495a08fe08dcfa5d3d6bd47f1684be6f6ad3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/metrics.json", "bytes": 2369, "sha256": "70cc274f24a4ce58dc8eb3d9c6dfa02cb3b85b9f6d420544f024fc132939d1cc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/facts.json", "bytes": 11936, "sha256": "3f04b677ae31f5dc676d6c00593b12417d2b7660cdf56f48e3755e3e7421bae3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/codex-inline-judge.json", "bytes": 2998, "sha256": "9305bc882cf0c324f48137be47ef8f307f56a6327e802c404e18aac8c9deb7cd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-20T22-13-51-668Z-p2w0/postmortem.md", "bytes": 7815, "sha256": "da0f5ab9fc49a8d9a500848db0d7107794a4b49145c07601967a023e49fdee44", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8d0d800bcd.json", "bytes": 3940, "sha256": "9a6dbb4f13cbb10cd5f659e1cd2e46344035f9ba71442481de3e588b5b2d54f2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8d0d800bcd-desktop.png", "bytes": 10912, "sha256": "5e0b9154e68b25464c875892ac297b363155055ba2dd2414197b44dc5a4e84b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8d0d800bcd-winner.png", "bytes": 13565, "sha256": "f9c2e59d63ba06094e55a470f1497786eef77f4af75b37eff4c7d9cd8cf582bc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8d0d800bcd-mobile.png", "bytes": 8233, "sha256": "d93845630f542ca16c28504e9ed2842fdcdebc15c6753d46241b0e656af41694", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/workspace/tic-tac-toe-game/index.html", "bytes": 4567, "sha256": "030d6355397385181c6c9706492f008b38d8c907f620457dbca9069211a0cfb0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/result.json", "bytes": 861, "sha256": "4113ffd8a4c4b372967e4e562f38f2f00b8558e844895af8bada69c6fcc31078", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/score.json", "bytes": 3088, "sha256": "9484c0d3d7311aac7da963021532f3bc3bf0024bf41c15d8027ee77273de4d38", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/metrics.json", "bytes": 2369, "sha256": "2369ff1ec6b9d2d79295601c1caa45ea7e671737e8585abb9c4f23c6a8f38c79", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/facts.json", "bytes": 11721, "sha256": "9eb4876d9059ea8c0644d43eb49132356f0a69ca6d6c11595847c3519ecdf85c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/codex-inline-judge.json", "bytes": 3012, "sha256": "6e3c02fac7d66abef7d24405e428984f54c0f3c6f19485da0fba5fe7105275c2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-20T22-22-48-783Z-b95o/postmortem.md", "bytes": 7614, "sha256": "f5573015719bc9337612d046d9be6aedb1c158582487f5263aca57be5d414a56", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d74277a4e7.json", "bytes": 4031, "sha256": "abcb6c54006cc32dd4a8664d2ca295d5b2684d9f84bc41fed3cb6efb2c98c948", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d74277a4e7-desktop.png", "bytes": 10643, "sha256": "06a3017d054de1052af41cd270a9df5278561983ee8877c4cf546c06ed8d0c45", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d74277a4e7-winner.png", "bytes": 12954, "sha256": "ee7006ec06b92112ef2ea41218e0dab1c2683086f327e9092c82ad52da5bc071", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d74277a4e7-mobile.png", "bytes": 8200, "sha256": "74734828cb49d1858501fcdf6a3f38b1c62487cd82644b45a335bb53035cb4c1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/workspace/pet-shop-website/index.html", "bytes": 3015, "sha256": "986d14f325241192fe705ab9e08d8e2704d4cbf73d5b01fa1d9407bb60e94c1f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/workspace/pet-shop-website/assets/logo.svg", "bytes": 477, "sha256": "4e1ec45a6ba22a9a176d886a31f1479dba1260ccf92d98d96c0b95882bfe1bfa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/workspace/pet-shop-website/index.html", "bytes": 3015, "sha256": "986d14f325241192fe705ab9e08d8e2704d4cbf73d5b01fa1d9407bb60e94c1f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/workspace/pet-shop-website/assets/logo.svg", "bytes": 477, "sha256": "4e1ec45a6ba22a9a176d886a31f1479dba1260ccf92d98d96c0b95882bfe1bfa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/result.json", "bytes": 866, "sha256": "8f2397e2cb665100c85ecd25a92831503b3fc8f0daff28ddfc8952683b10cd57", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/score.json", "bytes": 3669, "sha256": "738df324a2ee585fc10bd65d05b0fb8cce0e9bfd53a2256eec828816fb8045b4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/metrics.json", "bytes": 2384, "sha256": "1d17a6e7beebc2b710006866664b53323d4491c8b4af15ab81c8ffd297b9a8c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/facts.json", "bytes": 15005, "sha256": "a7937881d2d60a0362e210d1b193ef187fa6fbef6116e88b27785c0f4148b472", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/codex-inline-judge.json", "bytes": 3564, "sha256": "a12eb51704ef169bf0f6b214afe3ed21ebc0f78a8373df1ddaeece72ec72a8d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-20T22-25-20-904Z-aphi/postmortem.md", "bytes": 8143, "sha256": "2ed143493493653efdbdc14b7bd86bd50b93d5782379e426cfa62c5a757c8540", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d9806fb77b.json", "bytes": 4755, "sha256": "c986c0af41cc7995b1b53a0de37ee75a12c458acca93be97132dd43217cd877a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d9806fb77b-desktop.png", "bytes": 180924, "sha256": "6fe45911debca438fc454275738a127df00686e2b2a25711876f1fc971922d61", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d9806fb77b-mobile.png", "bytes": 153374, "sha256": "a763ba50f6514e5e43d770d585aa3d702f26873f6c989a63953a376eeb2100fa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/workspace/build-me-a-single-page-pet-shop-website/index.html", "bytes": 5337, "sha256": "58de5e25d7d7960299ec0067769dce1efb29e3d33c5ba84e2ed5aa352e119b24", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/workspace/build-me-a-single-page-pet-shop-website/styles.css", "bytes": 4846, "sha256": "fb0db303298d91471a10c61d635faf0d7f78e92096fa283c5db59d6fe5b9f104", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/workspace/build-me-a-single-page-pet-shop-website/script.js", "bytes": 489, "sha256": "92332ad6d9e434b9a4445d3011848c2f213ca475d1c7c98c53f77a6149ebc66c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/workspace/build-me-a-single-page-pet-shop-website/index.html", "bytes": 5337, "sha256": "58de5e25d7d7960299ec0067769dce1efb29e3d33c5ba84e2ed5aa352e119b24", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/workspace/build-me-a-single-page-pet-shop-website/styles.css", "bytes": 4846, "sha256": "fb0db303298d91471a10c61d635faf0d7f78e92096fa283c5db59d6fe5b9f104", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/workspace/build-me-a-single-page-pet-shop-website/script.js", "bytes": 489, "sha256": "92332ad6d9e434b9a4445d3011848c2f213ca475d1c7c98c53f77a6149ebc66c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/result.json", "bytes": 899, "sha256": "a4f74bfe27b768b610331b57fb755f69d60b2e8bcbe825d0c948074cb897bddf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/score.json", "bytes": 4065, "sha256": "6af970ecab38c7e4e27367d1d0ae9a0af1e9862adc1276ffbd8bfa627ac9d851", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/metrics.json", "bytes": 2376, "sha256": "12e513e01b61c79d85685ee0309848f596b34db3c92eaca69562a9ce59b08c6d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/facts.json", "bytes": 13898, "sha256": "033df733f0a2985ef152d4297074b662468991e2499d037b70286d38f08170e3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/codex-inline-judge.json", "bytes": 4407, "sha256": "c63a106b4ccde145bdbfcc39afb34adb213ca940693073d96d669f1448dc5084", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-20T22-35-00-985Z-q4jj/postmortem.md", "bytes": 8281, "sha256": "71f66f595558bb7c941aae2453c8f1560e208588d6e293d7ea20bb0e65b2d52e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5db999fd9b.json", "bytes": 5740, "sha256": "4909a7e2e76d98ee4b4a11067a9f4527650aeae198378da0ce42c4e7e2d69e2c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5db999fd9b-desktop.png", "bytes": 137693, "sha256": "34195f4dfdc54780159bec940c50fa615b845ba0e5f64411413feb95345c35a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5db999fd9b-mobile.png", "bytes": 138614, "sha256": "cfcef08baf66e71a5fd4acd75b3617356482dcf167cf626543ba84a1b100eb23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5db999fd9b-menu.png", "bytes": 141544, "sha256": "91a2ad575e1c8b6ac7cdc7c5da24305daf7e024a9fc3f61203c766e18df7d50b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/workspace/pet-shop-website/index.html", "bytes": 9363, "sha256": "ae961b895663a344ad36dec95818f6419149797297a9a5b235253f9254904c3d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/workspace/pet-shop-website/assets/logo.svg", "bytes": 1132, "sha256": "0a1deefa5db2123c8a5e3d956370ef8ecb4a781df47f281ab137afe1fb3ade79", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/workspace/pet-shop-website/index.html", "bytes": 9363, "sha256": "ae961b895663a344ad36dec95818f6419149797297a9a5b235253f9254904c3d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/workspace/pet-shop-website/assets/logo.svg", "bytes": 1132, "sha256": "0a1deefa5db2123c8a5e3d956370ef8ecb4a781df47f281ab137afe1fb3ade79", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/result.json", "bytes": 1088, "sha256": "0cf96a152b33c48af4913f89f0a09ecc4717f59efa39e92715e6525c05bab53a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/score.json", "bytes": 4419, "sha256": "9fefa9df2de6cfdfbf22e61bdd10f9e516c97ab29cf6e9dfaad7813682feb35e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/metrics.json", "bytes": 2353, "sha256": "6bab64886c08dd4f08a80099a5ef3107b51ef53b500e9e2c945769d85af16892", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/facts.json", "bytes": 14765, "sha256": "d3d225048ea46833668341988d84c3c4538ef5781f4f6c39c95358b08961a1f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/codex-inline-judge.json", "bytes": 3525, "sha256": "ba7b7dd67ab0d91961dc5146003efeb6d4398c5d1ef1cb17b06907819a63dd23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-20T22-42-13-844Z-sf4q/postmortem.md", "bytes": 10519, "sha256": "e3f247a8806b3915fe74072200a5b52025c4d4d3227aae7fc85f45bc284c417a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4803b2395c.json", "bytes": 3370, "sha256": "750f9d58256dc50cfeb92affa93442a2a9dc71333e6536390c08a970d5fd010c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4803b2395c-desktop.png", "bytes": 478258, "sha256": "b6cafd342e587e4c6f3d7575ae0f6e73f517d7349834633da0f1c4470ddbcb06", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4803b2395c-mobile.png", "bytes": 356842, "sha256": "c55a71984c1014f086d928aaba8b227319e8ef527c00945d78559cfa03bf1af6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/workspace/pet-shop-website/index.html", "bytes": 10864, "sha256": "0c1da43ece69f3bdc12406119daca5215e5c9d43b5949cc72ac2a1db54d82162", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/workspace/pet-shop-website/index.html", "bytes": 10864, "sha256": "0c1da43ece69f3bdc12406119daca5215e5c9d43b5949cc72ac2a1db54d82162", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/result.json", "bytes": 938, "sha256": "b8989ac9d4da54bcc4ea09853f54837dfab28e815ac2abfe89a6c230ecbda38b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/score.json", "bytes": 3600, "sha256": "b6bae01bafa5ba1f2de79018b7ecc6dc529f5c8b474c8e4ae41525eb8cd74448", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/metrics.json", "bytes": 2414, "sha256": "a455957b3a106d64aea638a9c4055c976dd05edfca8f79881c0e0441d0ee8dca", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/facts.json", "bytes": 14537, "sha256": "4e85b80237abdbbd211ca069e28e1e3d063f5a28c9d3dfb8966cc1902122b28e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/codex-inline-judge.json", "bytes": 3080, "sha256": "71c52447aab66a282acf0843c58e866bcc0ce9661e944570b5fadfdc04406168", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-20T22-49-49-528Z-kez3/postmortem.md", "bytes": 8212, "sha256": "7b40a99f531403a5f0b6c32a1a462139ad2ed57897cf03931eb618677dc0a75a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/0ab63f0f62.json", "bytes": 4200, "sha256": "03d9e43d7c82092cb2b9ad1444af663c6c5979bfe00d4a9b8631c50408d6872a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/0ab63f0f62-desktop.png", "bytes": 146264, "sha256": "593a5548ee330d1d3ca44ff6500cd2628ca8edbe167c7cc6db9453c5375ba66d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/0ab63f0f62-mobile.png", "bytes": 134165, "sha256": "470cc8e6d32eb7f17435762ec5c13eadfb2b3a6b68912aa62fa6f4f161b99f03", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-source-selection.txt", "bytes": 19977, "sha256": "1220cdc9b8f7a4ca3605b57cb05f10e331ce95ec0d51b3016c972336e33f4e9b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/result.json", "bytes": 1157, "sha256": "a420d9622f7fcac061e4daee048d3bb900bb42a67076e986b6d60ba3720ca52c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/score.json", "bytes": 4397, "sha256": "5faba92ad360ae7eb445e50e4be3ab83dad2c51f4bb057a3ddef5a54c2c35e38", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/metrics.json", "bytes": 2393, "sha256": "5551cf59e2192fd43e42616d9582acbf585ef7ed9bd5d8774b314b22787bed15", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/facts.json", "bytes": 13285, "sha256": "4b411cba1a72b32b3884e90b8bcf6d2a6104b241ad74a9716c15a546775baeb2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/codex-inline-judge.json", "bytes": 4261, "sha256": "1aaa3f7ca99365bd97afcd2cfb0246f7ba5bdf4797114db4fde33cf97ba5707f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-20T23-12-54-589Z-lyte/postmortem.md", "bytes": 11050, "sha256": "7fb5feeade19cebc288a3b18d9393ef44475e69eab63ae8f921bd7e29d8e5d09", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee.json", "bytes": 12480, "sha256": "e3648bd829c25ce8f68c70f7e1be79010550bc99db3b1d565ca91a4385a5a291", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-desktop.png", "bytes": 624923, "sha256": "28484e68971dcc6688e6d9caa443d3e7156b98d846a66f600a38f82016575ac3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-desktop-top.png", "bytes": 201715, "sha256": "554a628222b83e71a678058e466643df5a1d691c3a2a9162333cd563ddc0fa35", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-cart.png", "bytes": 167482, "sha256": "464679a7f27de7c2bbc32130787e945a980caf74fb5395c037565d63a3e51bab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-mobile.png", "bytes": 835170, "sha256": "1df3e3568134a82f30a604ac06bfee88e5bc87bb30109d8475a2079f2fe85d00", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-mobile-top.png", "bytes": 2743, "sha256": "7f25a72b146c75999955c6517fdb944ab82e100ef6ec38505edcb0f48bf57508", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-mobile-menu.png", "bytes": 42774, "sha256": "d7ac4852b0c7312fc45f91d5c02abcbf73d79227d697f84dab1702a5aefebf43", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-mobile-fresh-top.png", "bytes": 96728, "sha256": "583a899294ceb22d9d78dca943ccb7a8bbfc8f2f97e7daf1ac9b8be99665b30d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/5e42f6e4ee-mobile-fresh-menu.png", "bytes": 96728, "sha256": "583a899294ceb22d9d78dca943ccb7a8bbfc8f2f97e7daf1ac9b8be99665b30d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/workspace/pet-shop-website/index.html", "bytes": 7992, "sha256": "f287f99047c95c86f61c7a93fb42118f971b569c15a359bd3ca6c64831081052", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/workspace/pet-shop-website/index.html", "bytes": 7992, "sha256": "f287f99047c95c86f61c7a93fb42118f971b569c15a359bd3ca6c64831081052", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/result.json", "bytes": 1204, "sha256": "3125b112a454a5a095b53bef9c482ea4b46caa2b835c7b49729093aacc4c15ab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/score.json", "bytes": 4401, "sha256": "823213f5d9220f5f055fd5520dcee90b415da9ecac5b87720efaa90eb11034ba", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/metrics.json", "bytes": 2392, "sha256": "834b125b698ffbd8dfca10d77107161d17497003e247724f986627766ae6af51", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/facts.json", "bytes": 15560, "sha256": "60e356ea1c931d64b8042909af61698e67b47a4c653998b9ce0c9b5a6d4efc0a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/codex-inline-judge.json", "bytes": 2913, "sha256": "8dad62c032c097cf7736058b9ffc87e9c7ac1a76ab017ec6fe1c418885f2eef9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T00-57-41-802Z-lgjz/postmortem.md", "bytes": 10861, "sha256": "a21306a283413d1a9ab2606d3923519f8559c32008959ef5b1c1dc1ba067362a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bf0a2b6d90.json", "bytes": 3309, "sha256": "eed25474af76c79902ce71bcf45e331e18149d9fd157498903e0bb64185c869b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bf0a2b6d90-desktop.png", "bytes": 106200, "sha256": "ecc58e38f54b07735429f0d67e4ed7bd0a98b8feffba9fc5e902dcfc6679442c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bf0a2b6d90-mobile.png", "bytes": 103647, "sha256": "7d53ccafd33df9fc73047a4a5c88b73c91969833d069812bb9ec5cdbeb090e6d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/workspace/default/index.html", "bytes": 2724, "sha256": "65dbdf94590fe3b2ce50d048053c63348d7b63bb1d70e16253d963790100393f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/workspace/default/index.html", "bytes": 2724, "sha256": "65dbdf94590fe3b2ce50d048053c63348d7b63bb1d70e16253d963790100393f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/result.json", "bytes": 1234, "sha256": "81a5e267b3da38059975e47c0dab23cf3fd7d40e618b79eaa6cad52bc4a48fc2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/score.json", "bytes": 10089, "sha256": "1eacc7cd661cad2c08630598e548763bd420aea651afb9ec4fe6d27fb737e77d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/metrics.json", "bytes": 2410, "sha256": "0f5d6ae1c9e0546b6ce048c6896d9af129d940531e451cefc01c1e24c4819ac7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/facts.json", "bytes": 21831, "sha256": "715b91f88dee845a54e14c07c6b8630c69b5dd8a7242355688bb02a8e88f12d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/codex-inline-judge.json", "bytes": 3215, "sha256": "c268e76d728777de0f5731d14cfbb5c7f1bb9c2916611a7fa48bbadb9406c38e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T01-52-41-107Z-8mnc/postmortem.md", "bytes": 11044, "sha256": "1cda2bd7703221e74d15bededa9447beeff52088b41fed161414bfa718f9fd06", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/580190de45.json", "bytes": 2826, "sha256": "d27b346176d33182e6592e2ca1d8e664f7a85426eb69191530517586c4488f08", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/580190de45-desktop.png", "bytes": 58012, "sha256": "dd48daf5c82302b2e7e2574f0900684f65b4a4d7cd204e59bc34a9981a9cb313", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/580190de45-mobile.png", "bytes": 58073, "sha256": "0a209da7099b9d070985e6109dc7f1fbf1753155968b51811d1b530eddcb004e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/workspace/this-work-e/index.html", "bytes": 17102, "sha256": "429a6ba14ed9341076bea71e3d8465501fedf7f99505fe6fa9bb1a80b03125c9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/result.json", "bytes": 928, "sha256": "1541ad37e8cb6d666b8465228a6f0c4e142dacc80f35b9899291896ba00a5d5d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/score.json", "bytes": 3190, "sha256": "e32b0741c5439a73342a6329691184f6cc61f3e1f25c268e5b815a418c2cf8f7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/metrics.json", "bytes": 2378, "sha256": "911d4d4fd3f0958ea561cc83719879201348006a66e151f3800338cc2425157d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/facts.json", "bytes": 12962, "sha256": "22166947d004b30498cafaed3c75af251b84a48ace8a9639db88396eca73c90e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/codex-inline-judge.json", "bytes": 3462, "sha256": "46721eefa3b00deef4efa2c5ce27e56a064dce1733fd9da30dcaf6d08558954c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T05-04-50-383Z-l4tk/postmortem.md", "bytes": 8538, "sha256": "921aa697044f8877e484598e3219774e6e82ef029bf7e54654b04c13cab68b3e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/b8c3abcaae.json", "bytes": 18383, "sha256": "43377d347a017040f167a2c757b476a0c46e18cb83430676df4a69dfc74c7d1f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/b8c3abcaae-desktop.png", "bytes": 163022, "sha256": "3c13e56ec06f544e93e5eaa45e42d95b8c5b5b418863888c265faac7f2c4feff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/b8c3abcaae-after-input.png", "bytes": 156494, "sha256": "23c5972f832671c641ebd61bec9eab1ff7b298c31c8e0cab4e8ae0d1f5e2db1e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/b8c3abcaae-natural-play.png", "bytes": 167101, "sha256": "e19994b26740f16cc5628a2b3193ddff983c0a3d70b61756dd557c01cfc8e095", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/b8c3abcaae-mobile.png", "bytes": 37392, "sha256": "e8745bab1024e8ef05eec79002aad1c63129ceb7e5b95c62696e1cd9eea6468d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/workspace/tank-combat-arcade/index.html", "bytes": 10721, "sha256": "c9c50481ccea22ca8b4bc54c582932a50d871d48dfbf73b1b78a648fa8c68a87", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/result.json", "bytes": 922, "sha256": "6908c20eb9518591d64c23fef6dcfbbe37f91b6ee2c6cda6077ff5d58c67a800", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/score.json", "bytes": 3198, "sha256": "400aa3de52b25ab8239cac4bf7145fcb479fb9348899bca12d21c88bfccdf3c8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/metrics.json", "bytes": 2369, "sha256": "bedacb81134e185d3146448fc5ceca5640e712ae82c1bad123ccaf14f04a0853", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/facts.json", "bytes": 12254, "sha256": "21bb716b84ed917f2dcc0987dc0c233e6ad7f4126bea2d5684b397bca49491e9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/codex-inline-judge.json", "bytes": 3343, "sha256": "f564624bb758c9665f3308e046e975cce9c786fca1479e3cbf3c675b0fac10de", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T05-29-22-701Z-xrj7/postmortem.md", "bytes": 8598, "sha256": "70ae52e100225da59be673c16e77537e49a4e19776b19cca61f91e71c7f327bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bd466d8610.json", "bytes": 10110, "sha256": "daf2fe50ecc3c5c3e179ba6a0868c031f187cc456db7cfe9b2296e1cc1bf95bc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bd466d8610-desktop.png", "bytes": 8603, "sha256": "d253d38f203cc697b446c3cc3ccd958420b947e87ecc7cac393f972c6ce0af76", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bd466d8610-after-input.png", "bytes": 9480, "sha256": "bc12603cd2d4d6e54f956cb4bb4620715bcb0898bc1915c9ef53651e567dba19", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bd466d8610-natural-play.png", "bytes": 19335, "sha256": "a793c25c27a6a8d174737a24d3f74ac57213aeebecad20cdee52b538c894d13f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/bd466d8610-mobile.png", "bytes": 3902, "sha256": "81625e747cb36464f544b55c17e904a89f538ad68e6f82a8dcd4586a61da00ae", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/workspace/tank-combat-arcade/index.html", "bytes": 7813, "sha256": "b0c22d81fc661cd42e285437472ba3a071fcd8359f797e61f15f4f04dc3e1196", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/result.json", "bytes": 995, "sha256": "1ef2d1981a23d965c6428e856f02bb1275fc348e0e947e0b5ebbf6f1806b87a0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/score.json", "bytes": 3227, "sha256": "a8cb42d53c1d04ce209a4569595d3aaec66da74457ca027212602fb031ba4f07", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/metrics.json", "bytes": 2363, "sha256": "49e2a8da805fe4fe9256c131a372260b88e57228a1959be9e43f11bd5451aa8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/facts.json", "bytes": 13349, "sha256": "c5b9e9eb712274a2b30b766cecca9cd80dbedc8b4560c067a10a794491192ce0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/codex-inline-judge.json", "bytes": 3405, "sha256": "16f791c4d8cc5d6915034e1b2ddaa62ffc4edfa198a281f0b9135196a37a8049", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T05-32-57-532Z-cmur/postmortem.md", "bytes": 8537, "sha256": "c6dd24d12510113437393029d8762dc16407aba5dfd35516a433a1c0a0c5a977", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/e0ce0d6908.json", "bytes": 9592, "sha256": "74f11b8c642a4f2e9385585bdaa189bc549d794f30a3fee58e23d048609a1fde", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/e0ce0d6908-desktop.png", "bytes": 7404, "sha256": "5ac40c95902b4c7cef654089a88d59885f9edd27d28b65836bc07335a3696056", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/e0ce0d6908-after-input.png", "bytes": 7893, "sha256": "0b167aa3e6dcdc7235f4c42409fc0f6bcf3e6ccb985b10b14d5741158cc28406", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/e0ce0d6908-natural-play.png", "bytes": 13672, "sha256": "7ef4469a13c86ee44ece4a0ad2110a3926967b4dcd0a24bdc1a58209f85c20e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/e0ce0d6908-mobile.png", "bytes": 6283, "sha256": "1c00c1024c7a94155377cb00b84af9f4554aa2e3217f4930e2ef940011d92021", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/workspace/tank-combat-arcade/index.html", "bytes": 7852, "sha256": "e25fbda52aff0b5b2f0e777b1d7bb8be82fea8956921ca5bebc024f7e7caa9d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/result.json", "bytes": 931, "sha256": "c8083e78b9b60d02d22f39f15307cd65d0a462ddcfa5b32cccd031453d06dc07", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/score.json", "bytes": 3203, "sha256": "885fcaf411e9bb623213ab91601769d57379de0b246645414bbeee3bf74371c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/metrics.json", "bytes": 2379, "sha256": "65367f89be516af863ab608ee9d695e36e0c8ddd4a3d4e156ea2420bae7535da", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/facts.json", "bytes": 12545, "sha256": "b337620e72d537b17290b293874785511d58efef294a23d3012b781e7354d6cf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/codex-inline-judge.json", "bytes": 3301, "sha256": "6430121883091c1b93b5b89b1107f49ff5368e2bacb8eb7d82aed48c0c73d7df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T05-40-42-476Z-xppi/postmortem.md", "bytes": 8509, "sha256": "e2ad4888572963510e3731e73d8588a8846ae8c81e46c18cc7833a1f2304aec6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/036b15091a.json", "bytes": 8949, "sha256": "7ccdada0dd76fa727c66f60eccec8263a20e78e323c50243544eaf2dc4ee35ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/036b15091a-desktop.png", "bytes": 12520, "sha256": "1a72a25e09d23d54bc0cf61832871774a0ad035e694c4fab55a9fa7a66fa9618", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/036b15091a-after-input.png", "bytes": 12409, "sha256": "84164b3e2033eeb74e4ecca4c23d85aa156f3beac07df83b8083519776de5c26", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/036b15091a-natural-play.png", "bytes": 18778, "sha256": "83c09cf9927c65a21425e04ac2a4d13f0c3f8593d6b0f1203ae4a4811120c9f9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/036b15091a-mobile.png", "bytes": 14524, "sha256": "66d5dbb8970dd276097e1f6499ac6f988b279294c7bed35f15ef6597007fc3eb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/workspace/tank-combat-arcade/index.html", "bytes": 7759, "sha256": "5b547da9b4ca4687d2ac7f4ee2d89ea33dd64e37f7dff1718ffbda8bb3043e38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/result.json", "bytes": 925, "sha256": "13add0968a00900463e37b21af8111243e85660f3da3aea89e15d7a9fdf50592", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/score.json", "bytes": 3198, "sha256": "e3e3ce1799fbd26da7f078ddc15d9b72902f6fe2941c74dece15bf862c14ae86", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/metrics.json", "bytes": 2339, "sha256": "3de056b7b8ae00c95c9e168aba21831969caae807eeb6ca6cd1a2a44412bbb49", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/facts.json", "bytes": 12505, "sha256": "1d48ab44046e11f0ca17e14bf82164f85a841dda4f86a86ad1fc7e020398663c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/codex-inline-judge.json", "bytes": 3602, "sha256": "e877bae8fe6b35f76159cc17a6f1966988e61bd7ea21c73e79f84f9f55743ac8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T06-10-55-182Z-iez8/postmortem.md", "bytes": 8572, "sha256": "fabf2eef3a5011133c6e5176b54633f008ff40903b737d7c31a771d08610bc9f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1d7324a5a6.json", "bytes": 9221, "sha256": "16992466396960c64182bbff6f99bb69043c75dc16f757324e3238ac0aa10679", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1d7324a5a6-desktop.png", "bytes": 9199, "sha256": "409b625205bb961d734d4ddba5b3d8c263c3c602343c6eaeaabdbfc3d26c86bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1d7324a5a6-after-input.png", "bytes": 10490, "sha256": "950c7db3837d306fdafe540eb9a7790ca5ab75c3620ce2e7ae153f20f3da57d5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1d7324a5a6-natural-play.png", "bytes": 10707, "sha256": "58fcc8dca13f1aef57c1f10bf932357e376ea282ca31d97489432ad5491f5e68", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1d7324a5a6-mobile.png", "bytes": 7173, "sha256": "ca4f7d09c08f7906304fe1f4cd5ea26eb11e67d18f23a8e6103bec840d6d3849", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/workspace/tank-combat-arcade/index.html", "bytes": 9993, "sha256": "24f268087448e7b7d35b85384f2a5438a49f2ccbe5b2fde84601a3f815fe8a84", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/result.json", "bytes": 924, "sha256": "079c26c81584a931329501e5f90ae0eceb878fbd0518e1c22c469f1886711366", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/score.json", "bytes": 3198, "sha256": "4e357e926ee2c3f27ac2cfd9a5af9b8700f396f67259409cde59bdeefb9c1175", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/metrics.json", "bytes": 2370, "sha256": "bfa2e040566793c96aba50f29abc3bb57932c46ca252e899113b16806d5b165b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/facts.json", "bytes": 12601, "sha256": "772fa53dc913687015b48381385b82fae941f32c0ba4c55fe59352404a568c9c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/codex-inline-judge.json", "bytes": 3662, "sha256": "8562291ed2e3891017e7804c7184c5d3c70c01d7f4e2b236edab8b2f711d3041", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T06-20-45-367Z-xij0/postmortem.md", "bytes": 8516, "sha256": "6d619c7f7e622febfdaae81ce7e2d2af2220c407def9aa5f70ce12374992bac2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1b5d0b9657.json", "bytes": 13180, "sha256": "dedb9996a672efb796e560b5364c3e6d9d78a8174bd98d1f10953c77ae58d99d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1b5d0b9657-desktop.png", "bytes": 7595, "sha256": "b0bc99ff84d697f2a72d060d855a2335170fd5a6e8328815a6db96ba9885843b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1b5d0b9657-after-input.png", "bytes": 8049, "sha256": "4756f7a2c8df31a4085e4446e919db14ac243531f9bdd830e3279d29a3de8240", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1b5d0b9657-natural-play.png", "bytes": 7581, "sha256": "becfb37ec3a6c0dec7da000ae6f699608d42864b79bf65d8106d1d68d5bc0dbb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1b5d0b9657-mobile.png", "bytes": 5316, "sha256": "c05c6c1d7fe2330151f19f34effeb39652bb350f589d0f12171508fba942c9c9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/workspace/tank-combat-arcade/index.html", "bytes": 6822, "sha256": "82f7025a51893580c8d1f6d9fec66c1afe7e608825e9f4986b69dd42283f3b95", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/result.json", "bytes": 924, "sha256": "0db02a2450e2080b6d2947e702a58071f5937ac1dd7d2ff301248990022f3541", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/score.json", "bytes": 3198, "sha256": "9afd2ac851b156f09a4782bad86738ed9b9ce28eb35f6227ed77eb5d42bd0cff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/metrics.json", "bytes": 2332, "sha256": "36d4177f5fb627de696c0ef08124e5cdb5497dd6e52a75ffc991d5932da6f936", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/facts.json", "bytes": 13669, "sha256": "9cedffb340405798f8386ebe3405bd83e81c8134a6591d7d4b6c253e4d115a82", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/codex-inline-judge.json", "bytes": 3713, "sha256": "3ac683b1b9e3fb95ef622fa00afa22d785d9477efb9473cdd07598def86e150d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T06-27-15-601Z-maot/postmortem.md", "bytes": 8501, "sha256": "eafafa2f51227472daf92a3dcdb6468e85e5b014b2fa0deae931ab46311d1f41", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804.json", "bytes": 11040, "sha256": "2f50826ee217ffe09713473219d34790e09ad5ad4fb9830f8725e53aea2dd8d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804-desktop.png", "bytes": 7456, "sha256": "6c321ae7ec67a73d8275f4e65b9424a786391daa89d3dc3909587375e7964303", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804-after-input.png", "bytes": 9008, "sha256": "f04ac61994bf03bcbeae1cf0655afe1a3d368b6176e6f3dc96e7296037262e18", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804-natural-play.png", "bytes": 13318, "sha256": "7f48dfa7cbacb4be23bdb92f974f848a1cde573e977d681e398d277181741c63", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804-down-boundary.png", "bytes": 11942, "sha256": "1cbb283adba9eb67cd0cbd42a7caa1d5c767168ef1be221d457b91979ebc8dcb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d0a218b804-mobile.png", "bytes": 8894, "sha256": "499143cb0f22ffa10ba0a0c6848730be61f66180b3d94cb3593c469282a51fa1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "32ed67b8728ed6a20fd6434d95ee93e16921b68685d98cca60465f3b43414bd6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/handlers.ts", "bytes": 624, "sha256": "ea886137ca6d3028892fa1f67270b25f04fd425710cdd6ccdbda6249256d3e75", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/migrate.ts", "bytes": 353, "sha256": "4bbe9c935f17d19f386517d2fdb0742ecb7f6e173ed50f522e551329ef051046", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/tests/migrate.test.ts", "bytes": 1066, "sha256": "256f2c7c2badd60909337710c259a069b8782aad4a8e8079b85b499a87998163", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/MIGRATION.md", "bytes": 3901, "sha256": "ecd1fbb8e045ebd207e3e01a0cf1c30d3a540f7f7fc0df5e6925476ae6776730", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "32ed67b8728ed6a20fd6434d95ee93e16921b68685d98cca60465f3b43414bd6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/handlers.ts", "bytes": 624, "sha256": "ea886137ca6d3028892fa1f67270b25f04fd425710cdd6ccdbda6249256d3e75", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/src/migrate.ts", "bytes": 353, "sha256": "4bbe9c935f17d19f386517d2fdb0742ecb7f6e173ed50f522e551329ef051046", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/tests/migrate.test.ts", "bytes": 1066, "sha256": "256f2c7c2badd60909337710c259a069b8782aad4a8e8079b85b499a87998163", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/MIGRATION.md", "bytes": 3901, "sha256": "ecd1fbb8e045ebd207e3e01a0cf1c30d3a540f7f7fc0df5e6925476ae6776730", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/result.json", "bytes": 911, "sha256": "a501d98986565680b0d366f7311fa6cc1970693a7d5ae1d0700318dcb2a2ebb2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/score.json", "bytes": 2933, "sha256": "888c4fd5b43aa145210b2034b981d89e69912dd75fa036c3e2d5b295943f3601", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/metrics.json", "bytes": 2335, "sha256": "e61e607484ad159a0edca15160709461b2f382ed7dd480cf8162eb9877aa70bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/facts.json", "bytes": 16607, "sha256": "9ccf74a5d996e092baa6cf88d1f50c9b3107a8269d1ee750c1b953053419800a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/codex-inline-judge.json", "bytes": 8428, "sha256": "6eea57c830f25cb3d7a5faabcfe5c594635893948cda515d42dcf6a5c0eac91b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T06-40-18-885Z-bjzt/postmortem.md", "bytes": 8357, "sha256": "5639a6453b48f810b35475cfef48677689cfaa1750ee482c720ee8441905e5ea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/0ab542d5de.json", "bytes": 10270, "sha256": "bca411bc1d9d602842652c6e7a73b872e95d8acd4eb200c50ae592ef7dbcaf8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/types.ts", "bytes": 362, "sha256": "ab0c7c1964e365a0228f598ae347d40e8a0880aeafa4435ae5477dda5a38e69e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/store.ts", "bytes": 835, "sha256": "6c877c17a044a87df7ed7af1488eb1f879d841902fd183eba20260a8fe34c735", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/migrate.ts", "bytes": 424, "sha256": "12b1a0ecad440fb037d3803b537788519c9495c251c57ae755f9d0d56492f5e6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/tests/migrate.test.ts", "bytes": 1317, "sha256": "127e9894dc4e61e9c7d4418be4da7464db26ccc926866983e1df5208499d0d93", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/MIGRATION.md", "bytes": 2752, "sha256": "345882fcbc0fb6a4a3ce266be032de51bae5ce880cf44d60cef455a33b75ae74", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/types.ts", "bytes": 362, "sha256": "ab0c7c1964e365a0228f598ae347d40e8a0880aeafa4435ae5477dda5a38e69e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/store.ts", "bytes": 835, "sha256": "6c877c17a044a87df7ed7af1488eb1f879d841902fd183eba20260a8fe34c735", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/src/migrate.ts", "bytes": 424, "sha256": "12b1a0ecad440fb037d3803b537788519c9495c251c57ae755f9d0d56492f5e6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/tests/migrate.test.ts", "bytes": 1317, "sha256": "127e9894dc4e61e9c7d4418be4da7464db26ccc926866983e1df5208499d0d93", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/MIGRATION.md", "bytes": 2752, "sha256": "345882fcbc0fb6a4a3ce266be032de51bae5ce880cf44d60cef455a33b75ae74", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/result.json", "bytes": 984, "sha256": "7f6a599643a97693f2e5a78868fdee8d5f64997af2d2eab392352d6606e568f0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/score.json", "bytes": 2974, "sha256": "9fcf3d2b9ea688b353256fc7c7c267d166a8f1c96fbb3884b62a62cb4bd11f49", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/metrics.json", "bytes": 2408, "sha256": "0bd7123ceaa114a768dea9adb0e2ce6b39b15447997d8714d02f304d7eaf91b3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/facts.json", "bytes": 15638, "sha256": "6a5008e3f4d257202487534026a34b0995cbcb1c79e94fd2e67f1e1839db0e79", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/codex-inline-judge.json", "bytes": 9005, "sha256": "60bf7bc73343c6e86c80314ab1efb88a0e7a8b00b80a07d3c40014e69b93c962", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T06-48-39-810Z-vc78/postmortem.md", "bytes": 8446, "sha256": "94bba0c8bf0e76a921575f3350aa7e68c1ba0141580739b2c82fc6dbd59feafd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/0c366c4af3.json", "bytes": 10844, "sha256": "9dc7182a1ea5cb936281e225dabe1a011129c65d972868bbf8348f26c06bf33a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/types.ts", "bytes": 428, "sha256": "6538f0768b7f8ad38c6d09017c7a070488221f797dd12bff35a3977d6ee42214", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/store.ts", "bytes": 734, "sha256": "241c5258f2bf712c79b27b5871f8b44fd0581b753a3e3d63dd3038541c81203a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/handlers.ts", "bytes": 537, "sha256": "f65fda80a1706958d14b1f14a3ddd6fe9c639d1aab38eb7b184277ae60c5d56b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/migrate.ts", "bytes": 866, "sha256": "a5ce7a4e73ab45c3ec5ec4336d65f510d5d6a3e4ca41877c898ca3ec00fa77bb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/tests/migrate.test.ts", "bytes": 1148, "sha256": "af7f30ca20034c5b6f669a22abdfad38c6cb85e86a81ca0cfa574c09186f1257", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/MIGRATION.md", "bytes": 3273, "sha256": "2e3dfaedd44a6f5d7df95cf90704cbb664a1af674c7c08104121293255beb97b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/types.ts", "bytes": 428, "sha256": "6538f0768b7f8ad38c6d09017c7a070488221f797dd12bff35a3977d6ee42214", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/store.ts", "bytes": 734, "sha256": "241c5258f2bf712c79b27b5871f8b44fd0581b753a3e3d63dd3038541c81203a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/handlers.ts", "bytes": 537, "sha256": "f65fda80a1706958d14b1f14a3ddd6fe9c639d1aab38eb7b184277ae60c5d56b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/src/migrate.ts", "bytes": 866, "sha256": "a5ce7a4e73ab45c3ec5ec4336d65f510d5d6a3e4ca41877c898ca3ec00fa77bb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/tests/migrate.test.ts", "bytes": 1148, "sha256": "af7f30ca20034c5b6f669a22abdfad38c6cb85e86a81ca0cfa574c09186f1257", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/MIGRATION.md", "bytes": 3273, "sha256": "2e3dfaedd44a6f5d7df95cf90704cbb664a1af674c7c08104121293255beb97b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/result.json", "bytes": 919, "sha256": "fb73711e0623641e45ad62fe2e9bd8fc694dd9b909de511450d44a9a1bac224c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/score.json", "bytes": 2948, "sha256": "0dfaf054995c6ce9e243a84c68bc65182a5f2729e969cb164de2cf9aef3523fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/metrics.json", "bytes": 2381, "sha256": "a788bca4cd8efcf13d9fa627cf73ad013a9715ae5cc822e0ad7255dc7e2e05ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/facts.json", "bytes": 14753, "sha256": "8f2a6f8d6392aa9f4bafd8da2fc895a35c3090d6b2ce2f44e9613fafa19e3903", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/codex-inline-judge.json", "bytes": 8513, "sha256": "32ff59d612f7d18d6f0db0665d8024c1ae0001f9d09dec05b2eba701b2aa6466", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T07-13-26-233Z-y6ml/postmortem.md", "bytes": 8570, "sha256": "839c86ac0874316b9981df3cc3341c964a3fbc6d3840ad1aa63c5b30b55c903c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/45f9a0afd5.json", "bytes": 10977, "sha256": "a2ecdef9e39c241cebb7c7e4091fd6706700f4035a5aaa571760189c58767999", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/migrate.ts", "bytes": 351, "sha256": "f7b63c0184579a96f90225a394410204a1737f67b10b4365b676461dbcfc231c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/tests/migrate.test.ts", "bytes": 1051, "sha256": "516017a35ed26c4317561eecd4b3d376ec8ede65eb324d01c2877fc59466d18e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/MIGRATION.md", "bytes": 3556, "sha256": "f2f0ef075bd277e4aa41dde07e96159c21cc74eb09340eec809215803bcc3ec2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/src/migrate.ts", "bytes": 351, "sha256": "f7b63c0184579a96f90225a394410204a1737f67b10b4365b676461dbcfc231c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/tests/migrate.test.ts", "bytes": 1051, "sha256": "516017a35ed26c4317561eecd4b3d376ec8ede65eb324d01c2877fc59466d18e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/MIGRATION.md", "bytes": 3556, "sha256": "f2f0ef075bd277e4aa41dde07e96159c21cc74eb09340eec809215803bcc3ec2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/result.json", "bytes": 913, "sha256": "9eb7e98ecd7ec427f27c339b07edf23d421ff0391dfb304c23a152488359306b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/score.json", "bytes": 2945, "sha256": "9e75d577f3d646d6cd5c0748b30e1f879d98de924c22aa8f336d206f9cf2a63d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/metrics.json", "bytes": 2348, "sha256": "e7e918ac93e2a4dbcbc119b5e0d150b553d84f04451ddb86207474f2b5a153fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/facts.json", "bytes": 15858, "sha256": "2767a679529421c859cf69351080c3eb7b118385da12c83cc33bfee44db78bb0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/codex-inline-judge.json", "bytes": 8410, "sha256": "7ef337f660ae6b1d5d0d9a829b87739e0e84f7658ea423b2b3f2f29345a791ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T07-34-37-945Z-1nqc/postmortem.md", "bytes": 8564, "sha256": "51357f3cbc308c5b944f094114fde7d4529b20d20ef2ed04706d638ce38c45b5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/c9a0db57f3.json", "bytes": 10270, "sha256": "40b1832bf5088859cc27633ff727bbb32e0ddc0506a4b31820a55efe7ababcbb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/types.ts", "bytes": 193, "sha256": "0c9a266194352ea16117ff3b7dc5471f7548d9eca2b1f6dbe5c9d7564ba136ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/store.ts", "bytes": 1050, "sha256": "0496ac9414f3bf7e03c43928e5869b6abbff1ea1993015e78985afa2f702200c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/migrate.ts", "bytes": 798, "sha256": "6287c5fe63b71539bea13bc3e80f307925222320da2f91927109e5b49cd26dad", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/tests/migrate.test.ts", "bytes": 2281, "sha256": "d32f3bc462e5670ca9d3ae9ea4bee8dc040996d0eaea783e1a4a32965db5e7ff", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/MIGRATION.md", "bytes": 2597, "sha256": "e00565c71783556778ba6469098d95f6ea2e09f51c83925da1095a62d07bed9a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/types.ts", "bytes": 193, "sha256": "0c9a266194352ea16117ff3b7dc5471f7548d9eca2b1f6dbe5c9d7564ba136ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/store.ts", "bytes": 1050, "sha256": "0496ac9414f3bf7e03c43928e5869b6abbff1ea1993015e78985afa2f702200c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/src/migrate.ts", "bytes": 798, "sha256": "6287c5fe63b71539bea13bc3e80f307925222320da2f91927109e5b49cd26dad", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/tests/migrate.test.ts", "bytes": 2281, "sha256": "d32f3bc462e5670ca9d3ae9ea4bee8dc040996d0eaea783e1a4a32965db5e7ff", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/MIGRATION.md", "bytes": 2597, "sha256": "e00565c71783556778ba6469098d95f6ea2e09f51c83925da1095a62d07bed9a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/result.json", "bytes": 912, "sha256": "c37d798bab117fed5a894c7643b351da3dd9859f215a5ca87a63cbbf4a44ba44", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/score.json", "bytes": 2945, "sha256": "0777e90c2326853dbbf8ec9e561666254578dea4245c123d0ed14279bb482f24", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/metrics.json", "bytes": 2337, "sha256": "a06e85414b589db8b97c6b5b6973beda76680486c9bb43130ab06f4796ba1d0f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/facts.json", "bytes": 16297, "sha256": "f574cd0b9d3014c3184d298d429989e8df89ee14200245ff657064a717b3bff6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/codex-inline-judge.json", "bytes": 8651, "sha256": "a06917e8fe16b47088b78eb3a705d1797754e12efa8edbb140e140b84c63d8f3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T07-46-38-922Z-nhqc/postmortem.md", "bytes": 8590, "sha256": "66a2a9d2137f4a1a4a34365a0fbc2a4f81cdedc22c14de88b5b3d236fb33bd0c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/5290a21937.json", "bytes": 10272, "sha256": "e24611b4191bf8fdeebe701893c922f23b69d0c71cd9e79d80df43f530e0a0d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/types.ts", "bytes": 341, "sha256": "c855748fc4bb413fcb67ee299178d960967cf4b5b3ff8c69a344669b91d11aa6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/store.ts", "bytes": 1088, "sha256": "c8703cee25f73689571e8e104dff4de99539c1a1b2a7bccaf1fd94e485edc9fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/migrate.ts", "bytes": 267, "sha256": "10ce355b3bf3dc7dec01df79453cd59c17c5b26de8284c6c5ab5346155b2d152", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/tests/migrate.test.ts", "bytes": 1259, "sha256": "71e00b134cd32e4b738e46cb4eb784f8746428de768f11831c9e8cba4646b802", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/MIGRATION.md", "bytes": 2537, "sha256": "88779bda051f2ea38959094f60c8c0d7ebb15dfe4659b76716f6f7edae01e1b3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/types.ts", "bytes": 341, "sha256": "c855748fc4bb413fcb67ee299178d960967cf4b5b3ff8c69a344669b91d11aa6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/store.ts", "bytes": 1088, "sha256": "c8703cee25f73689571e8e104dff4de99539c1a1b2a7bccaf1fd94e485edc9fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/src/migrate.ts", "bytes": 267, "sha256": "10ce355b3bf3dc7dec01df79453cd59c17c5b26de8284c6c5ab5346155b2d152", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/tests/migrate.test.ts", "bytes": 1259, "sha256": "71e00b134cd32e4b738e46cb4eb784f8746428de768f11831c9e8cba4646b802", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/MIGRATION.md", "bytes": 2537, "sha256": "88779bda051f2ea38959094f60c8c0d7ebb15dfe4659b76716f6f7edae01e1b3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/result.json", "bytes": 912, "sha256": "e36841bb9ce7f61decd9c08558f00c5b67ee7e59ad4e28ae0c2a97c95c010fc8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/score.json", "bytes": 2920, "sha256": "addef0758b82b41e0254654ef9c34268f4d8003e6ba52375fc7e8ae0dce0a0a6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/metrics.json", "bytes": 2387, "sha256": "7ccf82ff81534929a977b5039138cf5f017f6acda6f62cdb04aadb47e31a70f3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/facts.json", "bytes": 19033, "sha256": "e65b701f38c5ea2b0ba1f5bf0cf14cd9ed33c1c8c288d464983504150a21a9ba", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/codex-inline-judge.json", "bytes": 8486, "sha256": "7daf8ab11c33f3da2fa916c89638fde8ec98d6d9c95f67d2f5cfcb7cbee176d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T07-56-04-459Z-zw29/postmortem.md", "bytes": 8502, "sha256": "dc9c7220ba4256815f81bf5cb0dba58961eeb3720519c78a589e2e47d67170b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/29bc00c4a6.json", "bytes": 10271, "sha256": "6018718bd9ba617aa6be32d2da9ba245493d0c1588929c189eec7a45b8f54655", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/types.ts", "bytes": 444, "sha256": "2f9ec7812733ad930b84faf7373b2f830c8154c1ad9429c17b210b74e0d205b0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/store.ts", "bytes": 792, "sha256": "c21141532351fc62a4192aaafd677fd6f8e7369367bda37e4c761fb0c2c5dc26", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/migrate.ts", "bytes": 489, "sha256": "0ba59963cfb9be8e2ea0967b8b4c01996156032c1bd0fda349b1f19855d7e03e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/tests/migrate.test.ts", "bytes": 1245, "sha256": "a374fc24a73b5bf8af0291c7127dd4edd8e87ea705aa10b9bcb0598335ea0433", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/MIGRATION.md", "bytes": 1215, "sha256": "48fd79439d91b1157651bd3e255a0971e70b91e0925a6a5d2a20b8ca62c325a9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/types.ts", "bytes": 444, "sha256": "2f9ec7812733ad930b84faf7373b2f830c8154c1ad9429c17b210b74e0d205b0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/store.ts", "bytes": 792, "sha256": "c21141532351fc62a4192aaafd677fd6f8e7369367bda37e4c761fb0c2c5dc26", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/src/migrate.ts", "bytes": 489, "sha256": "0ba59963cfb9be8e2ea0967b8b4c01996156032c1bd0fda349b1f19855d7e03e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/tests/migrate.test.ts", "bytes": 1245, "sha256": "a374fc24a73b5bf8af0291c7127dd4edd8e87ea705aa10b9bcb0598335ea0433", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/MIGRATION.md", "bytes": 1215, "sha256": "48fd79439d91b1157651bd3e255a0971e70b91e0925a6a5d2a20b8ca62c325a9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/result.json", "bytes": 923, "sha256": "95e39c3a32dcefe2218efe95164203f3015191eac6cd654b46f6de9c018acbf3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/score.json", "bytes": 2927, "sha256": "22d674ef6c5e9aaab129295ae5dae56f263bc2add8545cb0e6af692553e75dda", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/metrics.json", "bytes": 2347, "sha256": "29c7d487450789bb9b9ebf4a3ed01bd5236e35cfafbbeece3f39049d93fbfec0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/facts.json", "bytes": 15420, "sha256": "0b86d7c5d837bc2ffff05079aefa4246a29c7101ef874233453474e0e71f6b1b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/codex-inline-judge.json", "bytes": 8322, "sha256": "e40ee2133431afcb19709ac68ef8d8b2aa14e2bc774881aec5f727e89ff6bc1a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T08-09-19-816Z-wqrb/postmortem.md", "bytes": 8511, "sha256": "8841f3c85f695247c58d708668ad3841b34e0a5ec4e39114715246a90f5345c9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/58ec2d685e.json", "bytes": 10352, "sha256": "602e37ff7c43af8300da378dbb9bb20e1b8634f11d77462d2d77333ec73ba806", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/src/machine.ts", "bytes": 1720, "sha256": "0cb1f9b2ba510930311f3d108cef12d3538bec21e3bd853509767c459fe93470", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/src/machine.ts", "bytes": 1720, "sha256": "0cb1f9b2ba510930311f3d108cef12d3538bec21e3bd853509767c459fe93470", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/result.json", "bytes": 973, "sha256": "c78756a8d799906a0d8a741d2bf8ee53102ec176afe4043da448b9200f8a1a77", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/score.json", "bytes": 2881, "sha256": "6457165169a0e9f2a99030a57a87d80e71e51395f1e3e095564d39fe0aa068f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/metrics.json", "bytes": 2347, "sha256": "977241c0c549699b33d49693f0de1cd5f209b456271249d88e171051f371a78d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/facts.json", "bytes": 12181, "sha256": "f3be86733f58df08e26934209f17e676e28c36c67f0c3209b95b033db034cd0d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/codex-inline-judge.json", "bytes": 5448, "sha256": "574cbcb34a830c5763439d89db32c5a3b46b4f59f03576024a3073a32de4f20c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T08-40-19-114Z-745t/postmortem.md", "bytes": 8488, "sha256": "c1e44ee44e77e7fc660cc43394553e2c550ee79d0463be47f662217cc77c8314", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/984f57eb76.json", "bytes": 44850, "sha256": "ef1b3f9304e16174e092035f4db2116e57e7aa0a0cf00bb00e807154c9deed80", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/src/machine.ts", "bytes": 1816, "sha256": "046a068fd2669d3216bb0e53ca919f44ca38c5f4b13c090e64fd48936d57ed1a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/src/machine.ts", "bytes": 1816, "sha256": "046a068fd2669d3216bb0e53ca919f44ca38c5f4b13c090e64fd48936d57ed1a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/result.json", "bytes": 909, "sha256": "c047b8cfa1aba839f3fae981f684bf1e7c78e05bce18f8b4d12b8da9c250f2b2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/score.json", "bytes": 2864, "sha256": "a6e16440e2a0b85a76081a4215b008d626e35264b845045d919757bd8761e6c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/metrics.json", "bytes": 2374, "sha256": "c33f04814239c7a9579ab372b3331ac04a0c42ca8e4d8c3792389b128b5537fd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/facts.json", "bytes": 14019, "sha256": "6bd718f45765c1cd676cfaa010b662d211aec8195424a59ebb72a52263c0d4a1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/codex-inline-judge.json", "bytes": 5049, "sha256": "499408cd09e193a53859c3bc39dc108205deaaf75374ed723bd17dc92d58010e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T08-47-52-030Z-xpej/postmortem.md", "bytes": 8447, "sha256": "1d15610718aabb82fe505879f62e295ad169e6f93b84ba0df7358ba215352cfd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/7207f1d1ce.json", "bytes": 44603, "sha256": "8f1926036ef69a521b648f5fccab6df61e9594bffca3767fcac065c614c1a671", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/src/machine.ts", "bytes": 2159, "sha256": "b4ee15c23dcaae97025c8415106670400314eb077921119d3eda26be2c067c31", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/src/machine.ts", "bytes": 2159, "sha256": "b4ee15c23dcaae97025c8415106670400314eb077921119d3eda26be2c067c31", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/result.json", "bytes": 903, "sha256": "a0b400b6d8c0b231b9807335d83644fdaf1e198144940ccae66099ed2df6d602", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/score.json", "bytes": 2842, "sha256": "0b1f65caf0cc7cd47404654920554da0a2758eb99c335fdb469f644b2e4ce924", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/metrics.json", "bytes": 2322, "sha256": "17c231bf2ceb63ab6bc44c9a13d47144689b562e31e6b4e51bb22683182eca7f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/facts.json", "bytes": 11996, "sha256": "cc1c9a16618ec808eef33274fb6e8df3ff9646b4cd2733fd18c8867d3c982e03", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/codex-inline-judge.json", "bytes": 4921, "sha256": "561b743d2a9101248572f27928bf625c77f739fb1ec43792a1491afc0db4bb3d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T09-05-55-054Z-7zou/postmortem.md", "bytes": 8362, "sha256": "3c8742c574f9e381e25eb3a3c8059e9def5ad96fd20a3cf8d03f236a653fa4db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/1e1ef96120.json", "bytes": 45508, "sha256": "eaae2da68e2599c8f30a5ed0b16a01566b345f54a7ae25b1be41e765b43fd6a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/src/machine.ts", "bytes": 2359, "sha256": "2bc787a4b709558095c5fd6169e011d7202c52a89190d1f95a5cf4c4c874bc18", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/src/machine.ts", "bytes": 2359, "sha256": "2bc787a4b709558095c5fd6169e011d7202c52a89190d1f95a5cf4c4c874bc18", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/result.json", "bytes": 2696, "sha256": "8dbee18ba04e24745264aedd983ce7b6be808e83757e9e4798adfeb60ecda608", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/score.json", "bytes": 6247, "sha256": "3dfdbb8f98b95f8e369727be5e6982fffac0605c969959468593f913f0cd3521", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/metrics.json", "bytes": 2335, "sha256": "28a0452f4dea754fee20f098ddc72f07a30b145cad845ac9a9138c6f81294f88", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/facts.json", "bytes": 19501, "sha256": "a2f6388b106ae67b2e9191fb79f915dc1da47493a59f46c113f2af92f6614864", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/codex-inline-judge.json", "bytes": 5182, "sha256": "af5a3c59bed139db894b11d185bd7cbef00ca6521f1b1b2c61ae0269d04474f7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T09-08-44-902Z-jwbz/postmortem.md", "bytes": 10414, "sha256": "aeb74cffc561d1b5bbb9ecc3e982125b6b72b2f794c40dddbfe3a8ae4d8a2f75", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/aa491ce09f.json", "bytes": 28499, "sha256": "cf25e9c30668c4b05da53b8e28cc9ead7451677d88aa28fa63194b183602e6b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/src/machine.ts", "bytes": 1652, "sha256": "b7f34de19a94e6ce4252921d4e878f4a1e71e5cf9ed269d4a0c947c582470729", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/src/machine.ts", "bytes": 1652, "sha256": "b7f34de19a94e6ce4252921d4e878f4a1e71e5cf9ed269d4a0c947c582470729", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/result.json", "bytes": 902, "sha256": "3c1e3f8bf77152354a5854f3b361cfeaab8615daec9fa7d8925dbfbf1c375fdd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/score.json", "bytes": 2850, "sha256": "18bac6799738ed25d310bb7229e47310f40eab8e8fb5bd7e25d86b6a750f2ee8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/metrics.json", "bytes": 2374, "sha256": "b632f92ae1b8ea0875b0937cc1a571d0cf93faf9ec2a53e0a9c6f6fe129147fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/facts.json", "bytes": 15594, "sha256": "b02c724d1a4697508d0751ad0d12edca397ce40098769598dac8feb5505cd247", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/codex-inline-judge.json", "bytes": 4846, "sha256": "41834556cf4cd120be8e3a86fdab4929a96b9ac8e78024f5d7ca0f01a3197935", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T09-17-21-472Z-oq1w/postmortem.md", "bytes": 8635, "sha256": "308749ac395af5f3c255e8a8577e3a8c81048e3904f796b85cc4c9cfcb182150", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9a48f2a8dc.json", "bytes": 45508, "sha256": "a66d2e8367088fa7f31bf8f91343ee1147904768cb356412a1db7f05e34085f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/src/machine.ts", "bytes": 1326, "sha256": "c099c46dde1cb1cc248c2256dced666b4c8ce174b46fd5497955dec6af4beabb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/src/machine.ts", "bytes": 1326, "sha256": "c099c46dde1cb1cc248c2256dced666b4c8ce174b46fd5497955dec6af4beabb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/result.json", "bytes": 1530, "sha256": "1d463ae5fc65eab2534a0f4e1d3ed005171dabd37188d687f2ad24f225101afb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/score.json", "bytes": 4763, "sha256": "ca98856a393be24b01093da930a415830d6bf6beb9e974c9b9d14b24430c2e80", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/metrics.json", "bytes": 2382, "sha256": "d9787daf40be1c19aff50f07efb889d3f67ee7a5e7b24efd976d681ea10cebdf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/facts.json", "bytes": 18271, "sha256": "4220fca009101e043a165c5627f1144f143d707fc0039714060642817a5060a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/codex-inline-judge.json", "bytes": 5135, "sha256": "1a861035b5c08eebc73c6f99e6387d58877938704da4071b48804fecd1a8ccc3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T09-24-44-818Z-a1s3/postmortem.md", "bytes": 10411, "sha256": "ea851afe87ed1a0a5fc36899d8acb5868f52b3b8c81bfac92d591f30a1dc350a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/ea1c1d9f7e.json", "bytes": 27661, "sha256": "35deed3230a0c85c95a189366c48d950107561ba2263260b8ae9795c66a03342", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/src/machine.ts", "bytes": 2673, "sha256": "2de255df6800e9c2d280b18eaf099c392217bdb6ad69291ab9d225effd91d130", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/src/machine.ts", "bytes": 2673, "sha256": "2de255df6800e9c2d280b18eaf099c392217bdb6ad69291ab9d225effd91d130", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/result.json", "bytes": 1888, "sha256": "7ca2dcc8585b42f7cba00d853435f69d904de74502d417dbbe751b6938f2bf44", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/score.json", "bytes": 4975, "sha256": "f2f89a4554de12c835dfae1832b59a1a5d1283d230fb1ed1c024e51c914b90ee", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/metrics.json", "bytes": 2378, "sha256": "c5eac4d8c5c31f56421fdb3f5bffb32a40b7b7d1c44afe3070dfce3895f7c7d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/facts.json", "bytes": 18766, "sha256": "99921ff99285c1f0e32ae725666a204fb9cbf049e26eeac994eb048a6f55cb85", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/codex-inline-judge.json", "bytes": 5028, "sha256": "bf3efef59b7ceb65738a92c19f4ddbc0d74b20824e5371f4ae2429d5611c7b93", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T09-47-38-568Z-30ru/postmortem.md", "bytes": 10545, "sha256": "cc8dc79499151b4bad874794cd29b545fc8f40b1c4bd40af730ea266284fd70c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/f10301a18a.json", "bytes": 48105, "sha256": "73d2bbff3ab49a73d998598f933ec476c60c5f67c0145aa5cd455e61325a60b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/workspace/pagination-service/lib/paginate.mjs", "bytes": 1207, "sha256": "80e8260db0a7471db5227783d92b3b354ed2be453ca812c917b2b4570104eef8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/workspace/pagination-service/lib/paginate.mjs", "bytes": 1207, "sha256": "80e8260db0a7471db5227783d92b3b354ed2be453ca812c917b2b4570104eef8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/result.json", "bytes": 870, "sha256": "0d8c88af9326260f9efc8d1d857af63617ffccadcef80f46a833054dbea66daa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/score.json", "bytes": 2811, "sha256": "1a026055b6f85bef3e0f08495b273828d235cc217d2b9012e54d516a4aee6764", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/metrics.json", "bytes": 1309, "sha256": "61fa46194d79965c5add7af313280c9ca7e39e0944404ecd0647a79099b681c2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/facts.json", "bytes": 10592, "sha256": "0ab9f9b1e2735b493b5d2270d4679b077d2a278e17f13041ad70fdcd49f62d79", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/codex-inline-judge.json", "bytes": 3973, "sha256": "9951089319f62d5cf9db9473bcfe866b7bec1110e48fad81c1777a8027279132", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-21T09-57-57-494Z-1596/postmortem.md", "bytes": 8622, "sha256": "055602391ce192a399dd8a578bd7d886081a2ca1a430d9b09293773bba776810", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/aa24edd324.json", "bytes": 9664, "sha256": "97f1416be9a198fb9bec0f412f76cd6b93eeb3833752824a662cdeade6656b5c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/workspace/pagination-service/lib/paginate.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/workspace/pagination-service/lib/paginate.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/result.json", "bytes": 1024, "sha256": "39d6784bf62a48e23c9e8199c4e1769ca741bfa102c0fc90c8ab1a19714c92f5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/score.json", "bytes": 4404, "sha256": "4b8c0cc1536d0b7cb171db47a297151569c70fa1395494224f2f66e65873a609", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/metrics.json", "bytes": 1312, "sha256": "cdae8666049b9109fd5a8a8b5c600e5eba059b2c64b3aaaa0858ab9cb5b6428d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/facts.json", "bytes": 10913, "sha256": "b58920738f2aab20f2d6725c1eac63dea66ac8b1ed4a29cb29448c1c7e372dea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/codex-inline-judge.json", "bytes": 2975, "sha256": "e60109a2809f71934832011a99f1eca9680d1649da413e26ccb53a9af1ae9369", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T10-07-26-140Z-h1m0/postmortem.md", "bytes": 10037, "sha256": "9a0bb93925bd4183a2a308bc9bcc68d82e53dfad480aa789300b555f52303a73", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/7de98bb335.json", "bytes": 2062, "sha256": "a351751ff4d7fd164756227ebafb6569348cc3d308992d0e0b017ee137d3c561", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/result.json", "bytes": 864, "sha256": "74b64c7bf94a960398202f922320a8641efb812a860b6ad087c84d04b89dcc41", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/score.json", "bytes": 2808, "sha256": "43bca0ca11b5d9bde7e678279e8f38f87d04b8d223033150ad37799aa48b49a2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/metrics.json", "bytes": 1305, "sha256": "1232510e0f490d6561365a341ff4156d98360b2f6de5faffee3c50fc53087679", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/facts.json", "bytes": 10451, "sha256": "65fa5392d89c465f92afc90e5a066441e283fece186e0fb1d28eb02fc0b92bdf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/codex-inline-judge.json", "bytes": 3877, "sha256": "91bcba57c4412cb7c4d35412e12b1c4e95d4c5c2f62fac1de4d245b0b520e895", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T10-19-35-222Z-sxfl/postmortem.md", "bytes": 8635, "sha256": "4d4978a8b88d502d6d21a00373c15e54a65bd832c17d7060b62672afb4670c6b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9f5f1437ef.json", "bytes": 9649, "sha256": "a04cc1e45244e3864b075861b77a5027e5d1d9de2114f5d51ee6abaff7d90ffb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/workspace/pagination-service/lib/paginate.mjs", "bytes": 1183, "sha256": "7e3e16e293b6e0bb3e454b84916b6e81c66fff77f66a6d950251348b1e069837", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/workspace/pagination-service/lib/paginate.mjs", "bytes": 1183, "sha256": "7e3e16e293b6e0bb3e454b84916b6e81c66fff77f66a6d950251348b1e069837", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/result.json", "bytes": 864, "sha256": "2f7a6c89b6bff161e7daae9440a6fb9f57b4df8a49618aa817d55cb9f917be86", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/score.json", "bytes": 2809, "sha256": "7120accf37f6a98353d85e01dd0148622823216b997774e03a34caa4c28b5247", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/metrics.json", "bytes": 1305, "sha256": "b0880901e12f46a69ce4d4fbec29990680d4029a3846568dea2ba1ea9dc69af5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/facts.json", "bytes": 10454, "sha256": "3ccadf1831ad911ee1003e0cb7346df053bc02ec87ad7f2650c8c91f963e9634", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/codex-inline-judge.json", "bytes": 3824, "sha256": "197818bbf95322907d53b102e68eeac9b028261d39d5f2f04085a89867a735cf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T10-22-10-028Z-nlyl/postmortem.md", "bytes": 8550, "sha256": "ccda39cee3ca79770ff893a050ffb0ddc058b836340ea16ed85ba68c23d2dc75", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/c379c1f0b5.json", "bytes": 9649, "sha256": "b2b18f9ffa44646aeaca2ebee92809dae9b140b43576ebab9c2aad3990d26dda", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/workspace/pagination-service/lib/paginate.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/workspace/pagination-service/lib/paginate.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/result.json", "bytes": 1033, "sha256": "d1f8bef03e76f85e3eaa4559e6059e9ff035954ac8a158de2e5e8839d809023e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/score.json", "bytes": 4408, "sha256": "9d16d4c801f0c30d64828f56dc54ed59e6acf7068dda97c4ad36cf8bf75fc5fa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/metrics.json", "bytes": 1308, "sha256": "1fb4e574879ffe17dd1106909c0c576ef01a9dc89f4bd38a1e19053345054547", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/facts.json", "bytes": 10687, "sha256": "5a061bfc6d0c445197b6e173971f539387263f10caaa8419de37b354848aea99", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/codex-inline-judge.json", "bytes": 2878, "sha256": "0ca6a5fb05b3686b05e87a529e4e029cbe1740e384f294e182bbf84099e54297", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T10-29-34-945Z-6g1s/postmortem.md", "bytes": 10004, "sha256": "17e43d2c18232f0c6365a9d61ef7b27d422a8894d36f8220f027d66b36045cc9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/d30d12396c.json", "bytes": 2042, "sha256": "1cd11657ce1b7c801f47617d2c1ae5bb2e507ffe63e9a97b319ad1ae02efde9d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/result.json", "bytes": 845, "sha256": "f0e3f56e4794bb4507119dd73854eba07ae999be27cac4f20c5af072ca11ea63", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/score.json", "bytes": 2798, "sha256": "79ac32e20d0efe6b7f9a31a189e1c5f67358fec9d95bdff9ce47692b215561ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/metrics.json", "bytes": 1309, "sha256": "4c8dea747d545e1cfd8b6b25a6780fd16d5be525b1dc1f50bb325c0003348a65", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/facts.json", "bytes": 10701, "sha256": "187626acb6c1682815692156eb74140bc69e03788691733f73a6f46a380978c2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/codex-inline-judge.json", "bytes": 4006, "sha256": "752cd5d79bd392810b83ffaaa37a3eea3fc739d71a6698f32633097d3fa89dfa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T10-42-00-394Z-mhm0/postmortem.md", "bytes": 8577, "sha256": "07eeddae0dcf88ca8920b5da34dd2a486fec1944a62a1140f69b692ad1a9517e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/a40d87af32.json", "bytes": 9682, "sha256": "8e306e769b1dd49b50cf6f2c6bac615fa76180ae14256f934991673784c59496", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/workspace/pagination-service/lib/paginate.mjs", "bytes": 1206, "sha256": "c1d05f5017aa01e5589371a213e12c339805d2612a33f3151fe6829ad9e6c591", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/workspace/pagination-service/lib/paginate.mjs", "bytes": 1206, "sha256": "c1d05f5017aa01e5589371a213e12c339805d2612a33f3151fe6829ad9e6c591", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/result.json", "bytes": 935, "sha256": "e0187906946fb4836612f8d8f94c7e9b4132b0db4c1cb549eb3c5983a0bfe6ea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/score.json", "bytes": 2837, "sha256": "e62e519a9c3a959dbda969f0a5fea47acd529b0f3068af5d281058d1253b0a3a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/metrics.json", "bytes": 1309, "sha256": "8af7ce20b4930bfa5c7fc575ed98b85e9a0a7d0a42b5ff6e5c571139a8bb7cb0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/facts.json", "bytes": 10721, "sha256": "ea2c5389be29ae3ba8baaa8560616bf024bb33f86137045003b91bad020c90db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/codex-inline-judge.json", "bytes": 3899, "sha256": "bc9c992bfd9c7abed684e0bceab418740a4baff0fee832ea99ceb3c8123bc585", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T10-49-32-813Z-1lww/postmortem.md", "bytes": 8556, "sha256": "937335f888305b1da7a33355c880649b14f0dded8bf02979975e58ba6b2d467f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9535ee2140.json", "bytes": 9833, "sha256": "edf82c0c0bccef12164335275c30017c7eca46938e69ec86b972ef74b62c245a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/result.json", "bytes": 1015, "sha256": "f2553d2905b9a2cd6c642f5522e83d68e7ddfaea4b69bcfa74b54f34295fbdc3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/score.json", "bytes": 3574, "sha256": "956a810a722ba0c9c78b8d08474f7b0dbe289967cf83a7d799dba58d5766dec4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/metrics.json", "bytes": 1314, "sha256": "1f8e7d93dbec59e8fae38c1d6d5e8c252ec262511cd4af5f58d0ad379d0254b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/facts.json", "bytes": 10034, "sha256": "6570d45358e4059aff7fd8a65343c89d071e0bfd995177669132e8dcd78f91df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/codex-inline-judge.json", "bytes": 5098, "sha256": "613788fbfd664141efdb3df366dbdf1c2c86619d371dd07d01e96b561ae80dba", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-21T10-59-26-204Z-aijb/postmortem.md", "bytes": 9460, "sha256": "b019dd65fc3ed25723c569d24afbb8300402f7e094f6875dc1be061fff1ec860", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/5537df05ad.json", "bytes": 4015, "sha256": "89407235fbbbee009dfffebc0d74c13e39480697d91e27dac8e770f780cef070", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/out/customers.json", "bytes": 5389, "sha256": "3169d7ff892db5c1c09ec2ed32a68310708ed4e7cd927b16656813c50ce39d46", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 5582, "sha256": "ff19bfd346b7cf35b97d2ac89d0de6197c4344852ad3a4c01acf9adebf257f4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/out/customers.json", "bytes": 5389, "sha256": "3169d7ff892db5c1c09ec2ed32a68310708ed4e7cd927b16656813c50ce39d46", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 5582, "sha256": "ff19bfd346b7cf35b97d2ac89d0de6197c4344852ad3a4c01acf9adebf257f4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/result.json", "bytes": 1269, "sha256": "eae576516d6dc28d9b851f5b7acd42171f89fbe7d8c86fe49d38b23d26dc4ab8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/score.json", "bytes": 4295, "sha256": "12cb92e01bd744558b51f34562da502a28291644a40e632f6def4d2e125f785c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/metrics.json", "bytes": 2383, "sha256": "9622588f147fbdb906ee45691d53e1c9bbc52812a11b3d452f7d1882bc153185", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/facts.json", "bytes": 13754, "sha256": "35bcd700b32c9ddd255256fb3c72d4553147a4c30367572008c94f113e9e0ac7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/codex-inline-judge.json", "bytes": 6518, "sha256": "842576d378fb5b269f9509144177839e0c062bd365af843ae18ef5f85ba38eb6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-21T11-12-45-477Z-c7rl/postmortem.md", "bytes": 10444, "sha256": "a289987178c41f68b96b824761cbbc77e415e1eff61bb652cad956ac122eb0d3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/38b493a33f.json", "bytes": 14923, "sha256": "86361aa72d4b29dd368a779fdf6f09b52dbc9e8f6fe78d9645cb55ae6948f02c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/out/customers.json", "bytes": 7234, "sha256": "e053f10dfd8beba3207103d438389ca40a3c13c0e49d2cbf6c6c3b18f2f7f6bd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/tools/cleanup_customers.mjs", "bytes": 3018, "sha256": "48df5703b1991176d92ab05b984018a6497bbd66cba971a2ccf3623a47d99b1c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/out/customers.json", "bytes": 7234, "sha256": "e053f10dfd8beba3207103d438389ca40a3c13c0e49d2cbf6c6c3b18f2f7f6bd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/tools/cleanup_customers.mjs", "bytes": 3018, "sha256": "48df5703b1991176d92ab05b984018a6497bbd66cba971a2ccf3623a47d99b1c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/result.json", "bytes": 1206, "sha256": "0e67e5fcaa544b86d8c92b1e4b2cc57c2803392d2d0be7288f738e93c7ee7e7f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/score.json", "bytes": 4089, "sha256": "70ffa8efd5a05068334f7dc98d2b20c484cdf8a797b356e2903d826506f239c8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/metrics.json", "bytes": 1316, "sha256": "27e0d5711e5de2df72d21d7b184eaae107bde47ee33f59245b8c2c8903d3b7c0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/facts.json", "bytes": 11210, "sha256": "ce879d271ec891e4fb0a080c2f22c3f009f9a1f1ae95ada13ca02de874282029", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/codex-inline-judge.json", "bytes": 6343, "sha256": "bd1c0f8c612fb700aa8e61aa4aead44b3822ace29107a390884705e17baf000d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-21T11-50-13-631Z-e4wp/postmortem.md", "bytes": 10369, "sha256": "c6848a084710ebe743cbe3035584aafa61367d926b7625acd2d955cd195dbf29", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/00468ab8be.json", "bytes": 6687, "sha256": "9b50704da6bc7c342f00d61c80903b19e98cb7d7cd7354fe274c2d107192a5ec", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/result.json", "bytes": 1024, "sha256": "06dcbea7bbc5796d60bfb9cf1908cfcf796cce0dd9dd71f13f566f5f063ddfbf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/score.json", "bytes": 3577, "sha256": "d3bd88c24720edde01d13d858853c6206d029a2001abd4f6b982be0aa839f99e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/metrics.json", "bytes": 1308, "sha256": "9fa92987e92566b5cc233d2829ed18d7d8adb2544bff7b58c7db43e60550165f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/facts.json", "bytes": 9863, "sha256": "10719549e58da665f2499d3897563a8d7eeabc636590f6c22fc20301e9be4e30", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/codex-inline-judge.json", "bytes": 4815, "sha256": "d08873a819d5e52d92073e18e6742bbc23638c348efd3dc0cff793e149d7ed87", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-21T12-12-54-160Z-2mmd/postmortem.md", "bytes": 9383, "sha256": "13f9cd63dbb2f6658ccd4fd7ced19ccc8cbb1ce963948f9b6892c96a174f14d7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/b64dc50f6c.json", "bytes": 4049, "sha256": "c21098b6d5c0893e1e8c8ee37be5ad96b92bb4c4d46d790cb16872ecae0d656d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/out/customers.json", "bytes": 7114, "sha256": "0372ead54e49ba26e1ab2e89abef504b309058dad625033061cfdc778f65735b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3333, "sha256": "e351127a1e3ef416bdcf2423c7d2040342fac4af31ad8f4c8d02211cbaa46a6e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 3162, "sha256": "6b4e5043bd0f233a9d4a19f4c11ff8eec96f78fc797cf46f5bfb4f79b633e145", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/data-wrangle-seeded-checker.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/out/customers.json", "bytes": 7114, "sha256": "0372ead54e49ba26e1ab2e89abef504b309058dad625033061cfdc778f65735b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3333, "sha256": "e351127a1e3ef416bdcf2423c7d2040342fac4af31ad8f4c8d02211cbaa46a6e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 3162, "sha256": "6b4e5043bd0f233a9d4a19f4c11ff8eec96f78fc797cf46f5bfb4f79b633e145", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/data-wrangle-seeded-checker.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/result.json", "bytes": 1208, "sha256": "38f51f1da73198b6d51bdc7df7a47aff89d50e11ab139ba3cf7c1f21ecbc535d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/score.json", "bytes": 4060, "sha256": "d100c4d794f3aa5e72e364cab2e7e383487851339cf393c111fb8f60f5e8ace6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/metrics.json", "bytes": 2394, "sha256": "ece6162c94e2a58578295948bafa71cb36c9022d0ab1c82c8d8350bdfdee5a7b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/facts.json", "bytes": 14680, "sha256": "b14433239430295693ded1bf19f57e4cd81106a8e3e27cab4f511e0388ddb92c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/codex-inline-judge.json", "bytes": 7343, "sha256": "0ac4d73c894b98c1d77bb11ee38b8d48eeb5acbf02c595d0545be00f9bd7bb21", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-21T12-25-26-298Z-9cwy/postmortem.md", "bytes": 10299, "sha256": "c48fcbcc27d24c126f8f54b94a84501ffa7cc2624ecf13a13e44ceb7f50bf2b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/9b0de6ab8a.json", "bytes": 8841, "sha256": "ee37f3ef5853ea9f22694c710962312b5ef35a34e8e303299ff5a05bce13ed69", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/out/customers.json", "bytes": 5863, "sha256": "394774aefbedc637e0967f7c517057b872aa1c10459219de248c092aed2a4a91", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2541, "sha256": "bcd39323a837cb3cb82ae55d0547804a947ef6a5f76b4e6bd24f7c244d29f994", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/out/customers.json", "bytes": 5863, "sha256": "394774aefbedc637e0967f7c517057b872aa1c10459219de248c092aed2a4a91", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2541, "sha256": "bcd39323a837cb3cb82ae55d0547804a947ef6a5f76b4e6bd24f7c244d29f994", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/result.json", "bytes": 1283, "sha256": "3b43b8e8ef05c0e00d7eb09e8a97c0d0cec9b6ca39c0035ac853b468397a324b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/score.json", "bytes": 4285, "sha256": "f2a6431e74c02e245188f01de8b5ff0314fe371930aae9ead36051a5c3d878f5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/metrics.json", "bytes": 1318, "sha256": "9e34cfe5ad5fb38705be756a4ae2ca91a81259f2eb46bd5d0adeb5867abb0d73", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/facts.json", "bytes": 13376, "sha256": "bff727a995a078b10f8c29701a49a043e31a02a77c39ee7ddd69ae2233b6584e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/codex-inline-judge.json", "bytes": 7161, "sha256": "9dfff338e72ee2e386720ac92d0edf4ac80f0ece1be62f52ef97e61de39254b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-21T13-21-52-574Z-55u1/postmortem.md", "bytes": 10603, "sha256": "306ab5414b3cceb9fd57c5f3b9b8072bad96841cf5e011b08db878b65ec68794", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/87f441de06.json", "bytes": 6479, "sha256": "cb1fe84e9418630166110bb3594220efcf692912e5f0da539c0e4c602fdeac36", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "00db116c163fc65befcb23070aeba960e46c69cf92885bcf84f87a69bbd2f9ef", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3087, "sha256": "d3b8af040eada2ebf08d80765397ea9009cf3354490e563e3b243b83206c10be", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "00db116c163fc65befcb23070aeba960e46c69cf92885bcf84f87a69bbd2f9ef", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3087, "sha256": "d3b8af040eada2ebf08d80765397ea9009cf3354490e563e3b243b83206c10be", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/result.json", "bytes": 1259, "sha256": "0679fe463c22d44a5590ce73131d27476167f935f24fd60a1b5953905b990885", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/score.json", "bytes": 4401, "sha256": "edbabbee0186f9f630058388b9f0329bc0367fb99f2f1d18a75c974430c83aeb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/metrics.json", "bytes": 2340, "sha256": "827431395fc0dc5282ab3ee5cd5d0440593865f0c53ca0d451af2b562f04f1a7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/facts.json", "bytes": 12455, "sha256": "dd3c5bb049a279686617aebd174b7f796446854f3b2fb7c89a7f653e90602cf5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/codex-inline-judge.json", "bytes": 6465, "sha256": "8dfbfae1f8091b7e29d548312906883698e277902fd0f9db5040d627e1f2a315", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-21T14-05-46-455Z-o0kz/postmortem.md", "bytes": 10402, "sha256": "4800ee79fa6fb056fbf4fd9c3a8a8a324650046c1fc12c18abaef62d5764bf13", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/4c44488e89.json", "bytes": 6303, "sha256": "d10ebb3e294bb5461651ae99f908306e11630d904cd9299013105bae28e1d36e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6161, "sha256": "4deedd16c852376bd0976d9a1c64dce025986badf673ffd48d75e75ea5de5707", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6161, "sha256": "4deedd16c852376bd0976d9a1c64dce025986badf673ffd48d75e75ea5de5707", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/result.json", "bytes": 1017, "sha256": "6276b7566bb49ab6853dd16a766211914b2017e1f9b906f0d7d16acc065bd2be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/score.json", "bytes": 3056, "sha256": "2a1cc60c0293706ed0c1ca7e9c29d6dfde3ff56c798f3639cdf694286654ad55", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/metrics.json", "bytes": 2376, "sha256": "89a2a1c86f2b01d21e8b29f489652b60d98b325b777a05b7810498c940ec1106", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/facts.json", "bytes": 13539, "sha256": "6be692696466d4dd899540c9177f31dc85a966305095cd9a08cd1513e109db21", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/codex-inline-judge.json", "bytes": 6157, "sha256": "1d2c14397884f7c15e0e8fdcb8a86f5da00d1492ccd0b41451c63f17e15715e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-21T14-46-31-815Z-c6kw/postmortem.md", "bytes": 8383, "sha256": "446e93374e2372fe086bb3a7c1a8f9602b438826d5a597b6f1b1a0eba2f6f175", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/f6693144dc.json", "bytes": 6060, "sha256": "6f4bb09b8265cc1838cb39cafe86b27451d083c41ed7304c0c9780d2689a11fd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9247, "sha256": "1941a994a513a1bfbc5f7d66021a9c225d1883c220ff0dd6b7a734b57da8f386", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9247, "sha256": "1941a994a513a1bfbc5f7d66021a9c225d1883c220ff0dd6b7a734b57da8f386", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/result.json", "bytes": 1017, "sha256": "5eec9ba9dd6e0d92bfee5ff546a880a49861dcb02cd8543ed497c3529431d4bb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/score.json", "bytes": 3048, "sha256": "eb4fd43f120be65759d63cc388e6bace6708554268ee1b1d84c55898734ed329", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/metrics.json", "bytes": 2334, "sha256": "6044fc81a461960579d17617ad94360cc9eca9525c04722f3c2965acbbcfc509", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/facts.json", "bytes": 13074, "sha256": "16283c774234745a0555fc71ca3d85ad79ba9f63b1c66eb20724e6960058938f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/codex-inline-judge.json", "bytes": 6208, "sha256": "14b04468aa3afa7fefec166f37b3880ff4a40d97a076245da52f8173c29cfd3e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-21T14-49-12-027Z-r394/postmortem.md", "bytes": 8410, "sha256": "b63e60ed604b7d44904018e981821e12846d5e5f51c46e6fbc665ba7367c6fa8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fdfc6d5b21.json", "bytes": 6060, "sha256": "95375f5cd174d676300ccd688b8ca9d3fd85a16cae7f0a8c6e127e7eaa189e11", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8789, "sha256": "df8f8985b4f982bae88f19c70b6432c910e00709a90beed1e4a0d648f04f9bae", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8789, "sha256": "df8f8985b4f982bae88f19c70b6432c910e00709a90beed1e4a0d648f04f9bae", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/result.json", "bytes": 1027, "sha256": "52c8f6736d4b12a9e3a5a6921d93ab2f0aceb2bf0137dbc414462712983a5f1d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/score.json", "bytes": 3061, "sha256": "584a84b1b7417833dd3d41e907ba119cb7ac50ca4b88a871c4a04941b540d6e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/metrics.json", "bytes": 2376, "sha256": "ce7b4d81e47a40481a821caca903167a2d641ff2d22672133f514475a8f03af9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/facts.json", "bytes": 13013, "sha256": "67f380700cbaea5aede2dd120587744cff66e2e974a81ad5c4ef082333a74af5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/codex-inline-judge.json", "bytes": 6281, "sha256": "dfb4929b4dc618f6e6f3440ffe7bd60f050a6ed5de013b8c266ea4dab10f9570", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-21T14-53-27-429Z-nbgq/postmortem.md", "bytes": 8456, "sha256": "5cbae912865623d314d1a390e8c15aef83f9add425544b05787f545369622c83", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/1a025ebfa0.json", "bytes": 6114, "sha256": "aa576f8454972c1b9f05d120e6a9d58ba5dac0182b9ab2d4871a281f4870cc13", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6439, "sha256": "645b46cc83672400451bba1e8ef43bc53ae9ff85be7d41de6cd8708c7e79ec40", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6439, "sha256": "645b46cc83672400451bba1e8ef43bc53ae9ff85be7d41de6cd8708c7e79ec40", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/result.json", "bytes": 1016, "sha256": "2c8d611ddc2674c260f77128f3c577328687f833cd382db92836361177f469a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/score.json", "bytes": 3049, "sha256": "78a33c022f41988ef31ef928a7f222974dbbf462ca385f086c1bfc6f84571b2c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/metrics.json", "bytes": 2334, "sha256": "5072498842189b91531b904811d561c1a063e70ea98913f5fa62df8120492bcf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/facts.json", "bytes": 12817, "sha256": "9f7b75fad01de6677bcded24268e422868a76cf126240e3211264779f9dd52aa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/codex-inline-judge.json", "bytes": 6209, "sha256": "73a757394d6bc81af20c283c202908d1705c4746146b5563344774c214ed1ff6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-21T15-01-00-298Z-up1l/postmortem.md", "bytes": 8422, "sha256": "be65a207637e935d0ac24f3ae85cc2be3ccf6537b2707305c64b3ccfc8be8a94", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d0a4c22299.json", "bytes": 6060, "sha256": "f8f96f6b455afca2a8b0c84a61f4008153d7a1ab3da75bfd13eb865f260db0b4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8936, "sha256": "1a8c5ebb0e441f894be8d259d7608b3e9b52e1362102a36574145303c9d99aca", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8936, "sha256": "1a8c5ebb0e441f894be8d259d7608b3e9b52e1362102a36574145303c9d99aca", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/result.json", "bytes": 1088, "sha256": "18acb35d8d0828528092ffcef88ff047831b3fbc29b66a47b24dbe66d9a1c873", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/score.json", "bytes": 3077, "sha256": "5403c4a55869683e952ada7cfdb7874eb5cfdebf7d44f89ee865258642f9e93b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/metrics.json", "bytes": 1314, "sha256": "ad3446bc88c17f09ce13ad104d927b7803ba2ecd025fc53d284415987cd338bf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/facts.json", "bytes": 10950, "sha256": "2c45685401b1878eba2362c758c997a11c3327d9d27a6fb110ab056883706503", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/codex-inline-judge.json", "bytes": 6612, "sha256": "58e6741f4c9571734e868d46de2c0c7a89d7230e111736be2481d26334dd9cad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-21T15-03-57-774Z-mx0u/postmortem.md", "bytes": 8449, "sha256": "f3b974b0ef057c0e98cf72dec26fb9e7dd753f01ac5dba97f40bd069569c553d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/79046b4f8e.json", "bytes": 6438, "sha256": "79848bd574892e80233813e338e47b4aa8cd8fe1577f3f1f5def5dd2a1a92733", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9990, "sha256": "abec9146bc0b0a3e6864d66358d128de676b5ba08b12027ff6ea247a5818c986", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9990, "sha256": "abec9146bc0b0a3e6864d66358d128de676b5ba08b12027ff6ea247a5818c986", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/result.json", "bytes": 1177, "sha256": "5b8fd557e5703aceb0cc2572412b6c5e6a33236962bf4e9678b19d5a227581a2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/score.json", "bytes": 4219, "sha256": "edb910bf3f464563f9f0056c47de1671b7fd4d8f89d80e6264c374696017715f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/metrics.json", "bytes": 1314, "sha256": "abe4ab8e1b2347c68fc7e7cb6e95389f26d2239e4abf64214cd270085b1cd9f9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/facts.json", "bytes": 10993, "sha256": "4578fa5ef84a656966e4980af6880bf9e34df9760d2e35c0207f65e75210731f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/codex-inline-judge.json", "bytes": 6200, "sha256": "4acd29594d1d8c1abf1a8914ccf100793343fefbcad783777984aa12697672ab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-21T15-15-38-086Z-isza/postmortem.md", "bytes": 10213, "sha256": "208f2927f3afb2803cde54e22fcd5b443c9c8fbe90eaf8347a60e8630589d31d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/45668bd566.json", "bytes": 6087, "sha256": "01911ec24de1c515292894c56a5bef15de10ba658c2f21de8baaed4e72268da8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6434, "sha256": "44cfd66a4aa6d5a943d609a81a0daea77f5240705f3385774289875adc3fbbf6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 6434, "sha256": "44cfd66a4aa6d5a943d609a81a0daea77f5240705f3385774289875adc3fbbf6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/result.json", "bytes": 1018, "sha256": "4169f562d5deff34d5291c9fc268b7154ff8797a649dce7fd6fd47919e9b8b24", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/score.json", "bytes": 3047, "sha256": "f84a5c2502adb882c290cf2d0caab1c4eafecc0e878e1c1291602200e91bbcc6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/metrics.json", "bytes": 2343, "sha256": "0d20f455005cf7e64aa0b5e4d1a7e3bcaee658b4ecb4418e0c2fd30d7bf76ad3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/facts.json", "bytes": 12915, "sha256": "6c2aaf910bd4741d5493cdb94fdb350a8c842fd03211a0645d360fb88d704867", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/codex-inline-judge.json", "bytes": 6173, "sha256": "09fa38f50e3c9002223b2e0c9fc034208e7ad8a08c56ece25a77ef72b9b15e24", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-21T16-16-26-047Z-c4ck/postmortem.md", "bytes": 8411, "sha256": "92ff976431db868fed94b3aab804841c86657f1b17d17bdbcfb01c532b423a96", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fc6bcce7e8.json", "bytes": 6060, "sha256": "9daf19534c1d1561febe14910bc2bcbcc8256db0b0f21fcafcdece18b6730b41", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/runlog.md", "bytes": 642, "sha256": "be92abce16bdee649aaf9058b6479d5d5ffb6e63f3aa2f7d0b89b4e0d64249be", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 290, "sha256": "9689361c3d869293267fa026c8f455a40e154cc8602818446fa371798362e53c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d5145b9053-tools.jsonl", "bytes": 2220, "sha256": "280e8afa390deb99a55fd0948f2c40e82b01d4f887c3a5687000e1b64c1e167c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/runlog.md", "bytes": 642, "sha256": "be92abce16bdee649aaf9058b6479d5d5ffb6e63f3aa2f7d0b89b4e0d64249be", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 290, "sha256": "9689361c3d869293267fa026c8f455a40e154cc8602818446fa371798362e53c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d5145b9053-tools.jsonl", "bytes": 2220, "sha256": "280e8afa390deb99a55fd0948f2c40e82b01d4f887c3a5687000e1b64c1e167c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/result.json", "bytes": 994, "sha256": "0ff7d5e45c239417731705ca5c02905a9d9d9376f4ca25bda02004bfe94b5a3c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/score.json", "bytes": 2981, "sha256": "94de87bb8708b40db4cc39110f170fccfb74496611220dd51c673d975b6bd02d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/metrics.json", "bytes": 2328, "sha256": "158e693a02cc8fdd8a96e9442448565994c4efdb50fc81714c071cf9a201e5a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/facts.json", "bytes": 14145, "sha256": "da25cf40b727826ff3047f907aabc324e735be3cb5e4936a5c8bbf811f845729", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/codex-inline-judge.json", "bytes": 8793, "sha256": "0ff2f6a383e68477f44355b71a4c95a0d02e6d0aaf614c83e19273ce899d924b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-21T16-19-44-059Z-4hlj/postmortem.md", "bytes": 8325, "sha256": "a106138bebe0a5a861014e561e84d89870f9f70cbbe5f669f3eb5aaef51b7749", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d5145b9053.json", "bytes": 12189, "sha256": "d3da99afd5830fd87371ab5e5a448a9697e5ea02d23d97511d133eefa5f356a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/runlog.md", "bytes": 700, "sha256": "438dc851c97e9c8df877ffd4e923981eff2b56dd85be0db1de46edd22e15cd80", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1464, "sha256": "544f5a26cd854e5b5b187586c5adcc19bb59bfc98f236e2b6478535eca0ade99", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c9a224af82-tools.jsonl", "bytes": 1206, "sha256": "e2a0cffc7dcfb803be320d0cf9e76a20f0af0f75851568f7070e41caba991974", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/runlog.md", "bytes": 700, "sha256": "438dc851c97e9c8df877ffd4e923981eff2b56dd85be0db1de46edd22e15cd80", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1464, "sha256": "544f5a26cd854e5b5b187586c5adcc19bb59bfc98f236e2b6478535eca0ade99", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c9a224af82-tools.jsonl", "bytes": 1206, "sha256": "e2a0cffc7dcfb803be320d0cf9e76a20f0af0f75851568f7070e41caba991974", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/result.json", "bytes": 1004, "sha256": "e1bc0918b083d277dde725b5585b6a4b07cd6237bb16219cf2a01096361e5f7e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/score.json", "bytes": 2977, "sha256": "7c1b87985a5ee30340bb708ced06297e62b72f59ba89755d85c549bb81d5ea89", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/metrics.json", "bytes": 2371, "sha256": "d701785163f4e75e252369b2d3c2ffa48ad5139f0b178994a0b04d6266c6a029", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/facts.json", "bytes": 13028, "sha256": "42a6ca30120e3565766c72f31984e11a55875b7380ea65fce935e11ea4c54b05", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/codex-inline-judge.json", "bytes": 8864, "sha256": "017a143788f79b0822c69c06ab00f8ab7e77f2a2908a1d3fb673d80a48830a0a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-21T16-21-45-686Z-uifd/postmortem.md", "bytes": 8298, "sha256": "32717c37260a3d66346da4e7f6ad529fdc058b0c740a5622c2b8115772489ba2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c9a224af82.json", "bytes": 10866, "sha256": "add95ede895de6d2b2d6ccef6d01487c49e1eaad0fac261f10118164be2f5362", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/runlog.md", "bytes": 1076, "sha256": "cd3b6ac13c53d2f364e389235f64a95286bc77ecca86ac29d6d3689d5bc1890a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 459, "sha256": "f63b44306edc13017d4ce47e66e9101ec15741c2e00be8a874752ee02c2ed54c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fbc55aaa60-tools.jsonl", "bytes": 1438, "sha256": "7a3b17114656a6125481b0f23839497ae9e4d99eb39333f9f4af5f24681284e2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/runlog.md", "bytes": 1076, "sha256": "cd3b6ac13c53d2f364e389235f64a95286bc77ecca86ac29d6d3689d5bc1890a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 459, "sha256": "f63b44306edc13017d4ce47e66e9101ec15741c2e00be8a874752ee02c2ed54c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fbc55aaa60-tools.jsonl", "bytes": 1438, "sha256": "7a3b17114656a6125481b0f23839497ae9e4d99eb39333f9f4af5f24681284e2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/result.json", "bytes": 993, "sha256": "8fdee066d3c3c35cb3fc547ec91bb72f364c3224fc106f7ff4643743e4fecb07", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/score.json", "bytes": 2983, "sha256": "6a12ccbdb7bec46b4f10ca643b7c2ba4ea81a90ed5ac0e56d49ad442b053cc1c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/metrics.json", "bytes": 2328, "sha256": "69955b852016826262c4677c684743356807d15bbe6f6408ac88d62b52be1e7c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/facts.json", "bytes": 13010, "sha256": "79ff71fd27c6adde81142f392e2cdfcf4338d7155c809b7095e29065f6f28d44", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/codex-inline-judge.json", "bytes": 8815, "sha256": "4be605302bd01f29911f1a1ef2de4a20ecc19e2174958028f6ffbf5c94285c91", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-21T16-23-53-994Z-oy6c/postmortem.md", "bytes": 8346, "sha256": "3385469a8cac672f567e2a75d108316fa6f354e46834c7308a0b4048aa9c4467", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fbc55aaa60.json", "bytes": 11109, "sha256": "f2a9dafb9fbed09027a85e6239b85e2a85ee01466c5446c796154c7215dfdb68", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/runlog.md", "bytes": 650, "sha256": "8b153fdf585ee13d78363150bf6eac35c7463be533eca58a3c6305e557533654", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 837, "sha256": "1ed4f4f298c600aa7666bb70c903afdad1bad67f69839f85717848b6b06a4ee5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fc1a0352bd-tools.jsonl", "bytes": 1800, "sha256": "a7728daa76bb55900a39a64d40450c644e76ad99b7d1737d68792d6b73d6e846", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/runlog.md", "bytes": 650, "sha256": "8b153fdf585ee13d78363150bf6eac35c7463be533eca58a3c6305e557533654", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 837, "sha256": "1ed4f4f298c600aa7666bb70c903afdad1bad67f69839f85717848b6b06a4ee5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fc1a0352bd-tools.jsonl", "bytes": 1800, "sha256": "a7728daa76bb55900a39a64d40450c644e76ad99b7d1737d68792d6b73d6e846", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/result.json", "bytes": 1066, "sha256": "3ffa566b11a71fcfdc07b3c55d0a558eb19d547fb118e1940eef13335a80349e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/score.json", "bytes": 3010, "sha256": "eee4399a665159366cd9127c43b3eb1e9d76b50ae8b8fd29c29ecc11bb1fd492", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/metrics.json", "bytes": 2363, "sha256": "b593f5114dccdf234db0f7b8abb8f3c9636ec734e63cbde35e5e610d34d7f394", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/facts.json", "bytes": 14142, "sha256": "e8ced9271a446bc272719a48ab2b29286d0334a9bfec8bd0d60eef1fe46f7a70", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/codex-inline-judge.json", "bytes": 9187, "sha256": "7f65d761f6931fbcfb71d860d47d8a381147fa571ffe790b1091efef3ca2b831", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-21T16-25-40-745Z-bujk/postmortem.md", "bytes": 8348, "sha256": "a4f73ff22198904aa7333218cecfa0f9e7b2257cc03c73e6b2da6fced19f02d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/fc1a0352bd.json", "bytes": 12249, "sha256": "b191be961d3e7e552aa87dd07226da3143d0d7a18664ddb5370ec01b9925c1bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/runlog.md", "bytes": 202, "sha256": "837fc212343998c556a7d2617e43d73c035a88e65bec177641f835becfc3937c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 593, "sha256": "fbc589ecad64753c4ab4b7385768c9d8cbd38dec9d00f52276d365dc0d05555f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/33341c664d-tools.jsonl", "bytes": 1206, "sha256": "4e44d4d538db527accda94444a3f1d80d535982cc6d2027364044ec9b57de232", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/runlog.md", "bytes": 202, "sha256": "837fc212343998c556a7d2617e43d73c035a88e65bec177641f835becfc3937c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 593, "sha256": "fbc589ecad64753c4ab4b7385768c9d8cbd38dec9d00f52276d365dc0d05555f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/33341c664d-tools.jsonl", "bytes": 1206, "sha256": "4e44d4d538db527accda94444a3f1d80d535982cc6d2027364044ec9b57de232", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/result.json", "bytes": 1001, "sha256": "75adf3beaa03047aebfdcdca990a2a753de2c7d59efbfbbc3719811010faf76f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/score.json", "bytes": 2978, "sha256": "e357aedc88de5b77d67ad2d16e3ae75d95c2a0b73ce1e795f322e721307c0b6f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/metrics.json", "bytes": 2371, "sha256": "23611b860d320448c818bf5262054143e4e967ef937afd14e081dc3ff5b6baa5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/facts.json", "bytes": 13906, "sha256": "154d70cab36fca7e5cc0510fdb3a7d08ced0c964a50c4d2443218d7112c3c5fd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/codex-inline-judge.json", "bytes": 8586, "sha256": "bde6b0a11340d8c4512503d52eaef44e0abd2a512f68a999ce06c25eed735fe2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-21T16-31-00-674Z-phau/postmortem.md", "bytes": 8050, "sha256": "34fdb2d0f04dc871484163eec9a2499d4dea842670e9061ea173ad034bdac522", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/33341c664d.json", "bytes": 10820, "sha256": "eeaba1fba2f0b2857b889e6684826f1f80a73f20b120ae31ffd569c2ec89d417", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/runlog.md", "bytes": 106, "sha256": "0b278d6548c92e8b36c87754cefce5f5bd8bc206cad23f5ff36cf5a2e3e249a2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 299, "sha256": "40743a7b804f67a70e9c13f28f32b215c54e6ce141210112d2f42cd8cf87a08d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4a68e12831-tools.jsonl", "bytes": 2363, "sha256": "28938c58c3a3bfff9d01a7568029e8038f40b1fbcbfc7fd745e09ae447480968", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/runlog.md", "bytes": 106, "sha256": "0b278d6548c92e8b36c87754cefce5f5bd8bc206cad23f5ff36cf5a2e3e249a2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 299, "sha256": "40743a7b804f67a70e9c13f28f32b215c54e6ce141210112d2f42cd8cf87a08d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4a68e12831-tools.jsonl", "bytes": 2363, "sha256": "28938c58c3a3bfff9d01a7568029e8038f40b1fbcbfc7fd745e09ae447480968", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/result.json", "bytes": 995, "sha256": "0f7065cb7f7efc45a34c72089a445f11ada6a2bbf4ebb9c5b891b4ff5658edeb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/score.json", "bytes": 2981, "sha256": "0027e48ecdce3770a93c1dfd0116ff608397bdba33a299e8b6c877bfcc1bccfa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/metrics.json", "bytes": 2335, "sha256": "cdd56193da3c38a71d7a25c5b7f4a0a5805018acec351aa73d42a76f468dc1d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/facts.json", "bytes": 14083, "sha256": "5323cc4fd476be2f526567053c193a516bb7be5867cef4d05062535f2b68cd5d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/codex-inline-judge.json", "bytes": 8486, "sha256": "cf9be18fd088eb65430cdd9d5c01ade0cdca6adad3a42668cb67086b737d5e62", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-21T16-38-51-704Z-7oj3/postmortem.md", "bytes": 8265, "sha256": "10ba9e35fe70c3b4cba2fc4fd3b2e18a2f2bc44e74096a4689a22487efdaecc0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4a68e12831.json", "bytes": 12382, "sha256": "3447b357971a85d8b48422c6a603d6976bfa40cbcfe8c004e69c08b7c6214088", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/runlog.md", "bytes": 89, "sha256": "7fe339c5cd84d9516b1277123b46c27a3206f86b689f5c3d7f0b6e77a6cfc9b3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 519, "sha256": "5fb57c9c230518db3a646d832ab1c69b8226d7d93d7687d40ff8ee4a15c529c8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/25d6692d2c-tools.jsonl", "bytes": 2727, "sha256": "8ed0c092464007e61d29a4732dd8f87aca891ea85e8c00e70767ae558d6ca57b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/runlog.md", "bytes": 89, "sha256": "7fe339c5cd84d9516b1277123b46c27a3206f86b689f5c3d7f0b6e77a6cfc9b3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 519, "sha256": "5fb57c9c230518db3a646d832ab1c69b8226d7d93d7687d40ff8ee4a15c529c8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/25d6692d2c-tools.jsonl", "bytes": 2727, "sha256": "8ed0c092464007e61d29a4732dd8f87aca891ea85e8c00e70767ae558d6ca57b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/result.json", "bytes": 994, "sha256": "70e7d45f3343200e8ce55b0a05fa56cb422dd9ce3326fa6bf2bc0e0787d34ca3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/score.json", "bytes": 2997, "sha256": "eccdab17e7673cf3d208f77b9e21f59f412381bc16795484e17f4dcee5ae6a86", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/metrics.json", "bytes": 2367, "sha256": "78b0432fc13e874258bba4c207df70631ab84dd76eb5b891ffcc1ea72a11cf72", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/facts.json", "bytes": 14196, "sha256": "50bc4f362a4fde8bd3eb31570ba3e0f8545acbea3a4e972cce50ccb8172dafb9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/codex-inline-judge.json", "bytes": 8454, "sha256": "2656959ab594a33a8f297f6edfcb15324bbdebf78c5735562311063fe1025796", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-21T16-41-29-699Z-2wcl/postmortem.md", "bytes": 8288, "sha256": "5c4d93c2309a58afd08ee3f99a4b7bce835432f45d8286e522cb22a3b7c6e7db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/25d6692d2c.json", "bytes": 12894, "sha256": "dc4f9647a9862d08a2f53306232689c22606d77451191028fe2ea31d17e36f43", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/workspace/harbourview-office-relocation/plan.md", "bytes": 2913, "sha256": "a37803ceb36002185c2847e854fa8f22268646523f2b99f200b80e0ce5959b54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/workspace/harbourview-office-relocation/plan.md", "bytes": 2913, "sha256": "a37803ceb36002185c2847e854fa8f22268646523f2b99f200b80e0ce5959b54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/result.json", "bytes": 913, "sha256": "69274bac0292896ee2003a7aca0c19a1b525e5ffc20441e1d235189217a77e83", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/score.json", "bytes": 2907, "sha256": "c5001f7e7ce256ddc8cb9a7bc414ab356259ae0b7c6ac1df090fb1b360744572", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/metrics.json", "bytes": 2369, "sha256": "f2d082a9caf6553c298009619da1b32f9323699797f93a961c5e827b8398e307", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/facts.json", "bytes": 11257, "sha256": "aaa6005b18cfe58ecc9e52142dbc9473031917e508d9cf69724f1712781b405d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/codex-inline-judge.json", "bytes": 4168, "sha256": "cfe8352698f671627adcbc876a16aabeea7c9cff50e40ffca50242d615fed433", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-21T16-43-42-971Z-6jb5/postmortem.md", "bytes": 8335, "sha256": "18aec77537e4a63da9a510665ddd01e956ad27ca5d0bd37267a49853331d1845", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/40a973326b.json", "bytes": 3692, "sha256": "85e4bb9113d42179cd8814d96a2d4c65817209058a7a2818f222415f24ab8754", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/workspace/harbourview-office-relocation/plan.md", "bytes": 3444, "sha256": "78b5696bbac1308e6fb7fe642bc63d969f85b2328b121ceccc74d4329c9064dd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/workspace/harbourview-office-relocation/plan.md", "bytes": 3444, "sha256": "78b5696bbac1308e6fb7fe642bc63d969f85b2328b121ceccc74d4329c9064dd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/result.json", "bytes": 1358, "sha256": "fff48c745e8dcf80755c6625d7802e01c0db854290e374dfeec827d1573ed35c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/score.json", "bytes": 4144, "sha256": "90dba121fdc291b49b39686d6c408dc6463db5c20fc82260fe3e74b40ff567b4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/metrics.json", "bytes": 2340, "sha256": "dc4c4995cba45e133cbe787de563af32d8c89c8a782f5a6a0a9b948130b52d58", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/facts.json", "bytes": 13392, "sha256": "e2973a68972d21abed46f8c1f4050c6c8c1c54175abec8c64110d4472edc76bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/codex-inline-judge.json", "bytes": 4136, "sha256": "b5e24ea9e7dcfded987544c94ce6e4850a86340c7f191a74ba23600feab3049c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-21T16-45-34-573Z-pvny/postmortem.md", "bytes": 10083, "sha256": "ca9912b3eb86dfcf979b96477cdb53a65426f1df1607a49e45bf3af34d16926f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/56914e9354.json", "bytes": 3697, "sha256": "124f9b1b2e949edef700833d60cdf92fe6609cb55a5ef1607a1796bbbbca3c2b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/workspace/harbourview-office-relocation/plan.md", "bytes": 3369, "sha256": "4c5afaff8bd71703652686fbe37cec2dd5aeec67b5fe769502a9cdaf088c62c1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/workspace/harbourview-office-relocation/plan.md", "bytes": 3369, "sha256": "4c5afaff8bd71703652686fbe37cec2dd5aeec67b5fe769502a9cdaf088c62c1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/result.json", "bytes": 975, "sha256": "08dba2f649955a4f4c7abc7f5b56cd371ca5b46c393d03b8eb43a028ef30ae5d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/score.json", "bytes": 2933, "sha256": "a24fc430423d8863fdd9a7c5386f95c1fb78c7ab5f9d78c7b587a78c4660d590", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/metrics.json", "bytes": 1311, "sha256": "aa8aa9e2dd4da6b6408aac802833a2b48fe528a05ea762ee4e19881bb802b6e8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/facts.json", "bytes": 10358, "sha256": "12b609bab2290bc68a61e9adde9fe66800c68e72bc66858f35f3a533c344b97e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/codex-inline-judge.json", "bytes": 4367, "sha256": "428a8aed0d9ccafeb4dd8ca34f9285f7f983d30aae4cf1c9015af5357418c03b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-21T16-48-14-545Z-f7fr/postmortem.md", "bytes": 8370, "sha256": "8e8ccdf5737e0dddd7a904fa934db63166d24a4ecc120a0d44dbd3bc45153b80", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4d16f0ba9a.json", "bytes": 3896, "sha256": "858d2c5f9f7d063592f2912e6afa1f88234a8f49a5c2f5c568974a9adec0b01a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/workspace/harbourview-office-relocation/plan.md", "bytes": 4792, "sha256": "cf4f1584deb169149c13defae378c5d6d6f45eb54048363a8c6bb17355560df9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/workspace/harbourview-office-relocation/plan.md", "bytes": 4792, "sha256": "cf4f1584deb169149c13defae378c5d6d6f45eb54048363a8c6bb17355560df9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/result.json", "bytes": 910, "sha256": "9c21eaecda9f3ee24788a9fb26f98217c111703477ca052f8d1c4821e7c95e36", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/score.json", "bytes": 2907, "sha256": "4a55f92311b556dffccfeff0a82c878f74890b1b697623a08a949f63a6151c9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/metrics.json", "bytes": 1309, "sha256": "1e7e7976716980fb798fe691380dce4989d211aa4703a00c164bbf6d0f373291", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/facts.json", "bytes": 10264, "sha256": "7d4b57b34dd37d33cb21ba550865eef527627e07c43d7d682ed9628769b9a4d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/codex-inline-judge.json", "bytes": 4137, "sha256": "a1bbfe57897f65d3b51bde7f8ad2d07a0dc4166d72b4badba726a6c6deb5093a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-21T16-51-52-691Z-zplr/postmortem.md", "bytes": 8311, "sha256": "065e8369f325411ade9fb410df56527b76718f65d3f519b8372df23be23f38bb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5e1b2936ee.json", "bytes": 3675, "sha256": "662ec582aef3cee95192fc66b8c17e0816e208e359922739f2ca126424982357", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/workspace/harbourview-office-relocation/plan.md", "bytes": 2008, "sha256": "fd1695af4edb02be6032016d605baab0f81388e734a3a0adbd26d945e2ea9aa6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/workspace/harbourview-office-relocation/plan.md", "bytes": 2008, "sha256": "fd1695af4edb02be6032016d605baab0f81388e734a3a0adbd26d945e2ea9aa6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/result.json", "bytes": 904, "sha256": "9dac7c70ebd6bb2f07490d5505f348daac57ed8ed417de43807d0cc632591012", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/score.json", "bytes": 2903, "sha256": "99257030ad8ffc9672dd6c915a73d4bca8e9a0da3265d5ee9eb7b4bf96215a32", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/metrics.json", "bytes": 2364, "sha256": "46cfea6e9d46b01ce90eac393ec333fa0ec565cf2c07eba20dda9930fada3bad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/facts.json", "bytes": 11343, "sha256": "8fbd632fa792b1792426908b1ae4b4eef1d1d3978e15e4cca54e8cec9a80fe8e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/codex-inline-judge.json", "bytes": 4154, "sha256": "07bd1247675a2bf28a235654ac8942ea52b7ca0956077c3f52d2f91b3d43a006", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-21T17-00-32-491Z-ojwz/postmortem.md", "bytes": 8372, "sha256": "7fb1fd113b19d0afc66215e3b3e9a6f96d994a12a57a0f630d8031fff84e03e9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/8357a6a519.json", "bytes": 3658, "sha256": "4ad036310afb47fb1361218712efaf118d669defd91b9c6bfd49a4e2a81106ab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/workspace/harbourview-office-relocation/plan.md", "bytes": 2629, "sha256": "72cacfc35e6e986341ebd7d749d3d7ddac20d7e055456a3bad3b4645c8bf6b73", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/workspace/harbourview-office-relocation/plan.md", "bytes": 2629, "sha256": "72cacfc35e6e986341ebd7d749d3d7ddac20d7e055456a3bad3b4645c8bf6b73", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/result.json", "bytes": 903, "sha256": "4fe2988d5f68cf88d70831935358399d88d035570f239ac14a9ebb684ebc80d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/score.json", "bytes": 2891, "sha256": "5d653b060b8fb673769342dfe8f349569d5fc47143428a392fab46241938ebe3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/metrics.json", "bytes": 2356, "sha256": "8b31df5ab38e2f390d33a1ed9999bcc583999a658551fbe2b825b9c6699307b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/facts.json", "bytes": 10973, "sha256": "fa0eddee558f819c7240d161c7e8c93fbe00ce411ac616ec1f8467d762648988", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/codex-inline-judge.json", "bytes": 4138, "sha256": "012fecd0b234b999c722435e5510686dde36f79a671723e1a9e772fe837c21fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-21T17-02-02-683Z-s1ls/postmortem.md", "bytes": 8306, "sha256": "a47d7252efafa88c28abbc1377464aed6db20866a9f3712f336aeb5ff6265a5b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/381910c703.json", "bytes": 3658, "sha256": "5af3d7b0a611fab5b52455117a9f173779ee5cf230fe0f3c4b9975398bbdfe77", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/workspace/harbourview-office-relocation/plan.md", "bytes": 2935, "sha256": "6ae5bbbd554d66a1d3732f73b735b7516b3dcdd84264e5b17c24b553718bb9d5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/workspace/harbourview-office-relocation/plan.md", "bytes": 2935, "sha256": "6ae5bbbd554d66a1d3732f73b735b7516b3dcdd84264e5b17c24b553718bb9d5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/result.json", "bytes": 903, "sha256": "c3ef0dc7dc9d756139d2bc4fadf5c0d5e0cbd406de5b13240962da91b2761ca4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/score.json", "bytes": 2891, "sha256": "5d5c48006b0efd1c0712a2e907fe262e87655622513a8fdd79acc31de185d5b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/metrics.json", "bytes": 2360, "sha256": "b19ceab621f7ff77dec8b2e118bc091de88aad078fe9eac902235b4bb8b00a1d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/facts.json", "bytes": 11211, "sha256": "ca1ad79a0e423b99603f3f879667c1015dbab5d569f4b8415d912971bb1ea92a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/codex-inline-judge.json", "bytes": 4148, "sha256": "fab3bf71d8a4cbd30a1011dfa152ac5ee71e7215fa56907ab97e64e058155b9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-21T17-03-23-825Z-nm7x/postmortem.md", "bytes": 8326, "sha256": "a08d39b79239837dbd79dbd277f0203d34b2815c5d70d29649db38421b7a39e9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/1d4a5b87af.json", "bytes": 3658, "sha256": "8dd71108bb5c87e0a20d700ca7d417005daddf71b42eb89c487cbe9baa9109fa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1706, "sha256": "4bba40b835fda8ce5ee1106cca724cbf2d6ae46158beac650f7c582976c14eda", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1706, "sha256": "4bba40b835fda8ce5ee1106cca724cbf2d6ae46158beac650f7c582976c14eda", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/result.json", "bytes": 2685, "sha256": "373a2a6c2e3a0aa2251a5c7a2d83978a42e4464235f57eb2e0108f3f70ef4f03", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/score.json", "bytes": 6875, "sha256": "916f7d481e58daa19d67f1075bde25b2b204e70936624becf95da3ac6ed7191a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/metrics.json", "bytes": 2373, "sha256": "d7ada69835dabd33a46630a49674d70f22d142c9bca28d8f29902f514b88ba38", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/facts.json", "bytes": 19344, "sha256": "9908dfcc18642273a35f6a7eb72ef4e82727cdb4942fcb322146f712b4d0b4bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/codex-inline-judge.json", "bytes": 6712, "sha256": "2d9d3296df7619d52446e0c263f46361c18f3bac4b38c47e5a5b206fa488e1e6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-21T17-05-03-060Z-an5p/postmortem.md", "bytes": 10206, "sha256": "7cd9da12befbc3632522d36e3467733774f3cbf01a70c42d9d65b56a89d90964", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7babcba5c5.json", "bytes": 5954, "sha256": "57c48c1a1500c1a2c1d4731b92238802b1f923e51a9fa04e7356b5ebec4d4186", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2833, "sha256": "8e0beffb4580619920db594c4b1a9e45f654047efa24ce00ad5751a0ecaf97d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2833, "sha256": "8e0beffb4580619920db594c4b1a9e45f654047efa24ce00ad5751a0ecaf97d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/result.json", "bytes": 1246, "sha256": "26e1ed860687c56967a1506ef8e5e745c9564b450dfb17695b437d867fb4df25", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/score.json", "bytes": 3331, "sha256": "8f560c16fd11ed655d80422ec21ea8b806cc7e09eefc22167d3d12c4461745c8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/metrics.json", "bytes": 2372, "sha256": "bf5226b356b90310825c00df60c22d0e46b800c34f66e98fd9c12874bb990ffd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/facts.json", "bytes": 17303, "sha256": "398839213bce1ae6883583bd8c0fe0017291bb2a48257e3cb8d82afb238d15fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/codex-inline-judge.json", "bytes": 7140, "sha256": "43e9dbd4396470268b1ef412c91f4307372ee816c60fa08fb06a173625b6f516", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-21T17-07-26-611Z-glmk/postmortem.md", "bytes": 8477, "sha256": "383767fd177ce16a1c4e9c0ee7f1d0ea852efe9329808c694d865ff5d644a9b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/3c4a7c3b12.json", "bytes": 6416, "sha256": "1e42ba3b9bef954520f55af7de24b153cf816827eaa8ca06db0076ff0cf48ab0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2871, "sha256": "15999d09359e67bb63639eecf976360c89cc1cc618f48e0f1b91e8de0bdb1ff2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2871, "sha256": "15999d09359e67bb63639eecf976360c89cc1cc618f48e0f1b91e8de0bdb1ff2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/result.json", "bytes": 1181, "sha256": "576225eeca16c9b33dcfd7e689bebc3cac536ed77584682898977d66e8a04b4e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/score.json", "bytes": 3289, "sha256": "10c4560a7f28ad4c15604860aa9ad4f5367f50bd579ae17a9e2e9db300e9d4e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/metrics.json", "bytes": 2375, "sha256": "f574f37c7a91d0ea4aa77ad046b46290509ba24b3e566bf79d1147d85df80d1e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/facts.json", "bytes": 15723, "sha256": "1dfc8659a45035eef7e9c78b8c3830b79eef54ebdd3d21e43cc09982dc404258", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/codex-inline-judge.json", "bytes": 6589, "sha256": "a528f766d98991aa90a2eacf0086a70ab082de97b2085671a17e4497aaa52734", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-21T17-19-42-362Z-qpp3/postmortem.md", "bytes": 8509, "sha256": "4c5119ac1bae71e876ca88f4dbb5431acc295062751e537cc94da6e52f966b68", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c8b2a443cb.json", "bytes": 5987, "sha256": "ecaa47d8be1101b3c30b48a8cf81998dacbdc0b613e7d6b38ec9d9f7dabdfa41", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1277, "sha256": "2eb09681c3b2e9824c966cd25dcee02ca90265be79715cf63019d7b8a78aae30", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1277, "sha256": "2eb09681c3b2e9824c966cd25dcee02ca90265be79715cf63019d7b8a78aae30", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/result.json", "bytes": 1176, "sha256": "68789558bba50bcdbc57644153902d801c8fa7ea887be31e7566374f5ae5b380", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/score.json", "bytes": 3290, "sha256": "a26008b844a4587f55b5e22981e11c0f5e9a1228d7939eab062efd84bf00850c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/metrics.json", "bytes": 2375, "sha256": "6a69438a4cfbc5d3defea55903c3528b36972653f75f2db059e1e49976970d91", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/facts.json", "bytes": 14853, "sha256": "1d96461687a4f06e75fd4747df8ae72ef8cbd9da62a44e03730d48d3479f37f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/codex-inline-judge.json", "bytes": 6546, "sha256": "2527b5f77ff83ec1d4c1230135270eb1cb9f740a8697370b5ee7e107ae30621d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-21T17-33-49-266Z-ztf2/postmortem.md", "bytes": 8537, "sha256": "050e8ad8c1e1a00be46c49eb85b4e637904e7a14670500807bd657ecac8e2b40", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/a9a1313f0e.json", "bytes": 5954, "sha256": "fc6d2227e2090db10d7fee0cca204b82d06b2bbcfbc25786753ce0ca5b92ece9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1983, "sha256": "bbfbb41c8fc27bccf4d48da0f40bd0558e0e841988a3e78992dd51b3249a8de5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1983, "sha256": "bbfbb41c8fc27bccf4d48da0f40bd0558e0e841988a3e78992dd51b3249a8de5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/result.json", "bytes": 1175, "sha256": "45193e61ac967048870e19f6f27334986da95fd440cfc59ed55df30e682d000f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/score.json", "bytes": 4553, "sha256": "e293b37a5e8e886c79244e80017f7ecfd4e5530375768566a11aec9f1f5755fa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/metrics.json", "bytes": 2378, "sha256": "636d64d4a4be9c91b5c05797648d231d1118c054550bf6b00fa3a2f695e51a63", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/facts.json", "bytes": 16028, "sha256": "200de79d8de1ee529039b999b921f3ccca85999153c9944dd84e4b6b072f3964", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/codex-inline-judge.json", "bytes": 6578, "sha256": "8293c921e652295079a28f56684bdde0be57983e4d72e33fcbd6cfdacd39d45c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-21T17-37-34-142Z-n2o5/postmortem.md", "bytes": 9150, "sha256": "129e9cc17156715d37a63b235045414db10987cbb54b32c69ef4c11de1715c06", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5ac04a5fc1.json", "bytes": 5954, "sha256": "0f79c4eaa4c27dbd4ac8016fd620026c22404f9524d3963b5f0e86ff22784d18", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2336, "sha256": "20c19adc506fc8bb9741904d57308ac22fed8e74ffc002b9d50434dc3f97e17a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2336, "sha256": "20c19adc506fc8bb9741904d57308ac22fed8e74ffc002b9d50434dc3f97e17a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/result.json", "bytes": 1627, "sha256": "7900a5c30f309ac3e2af5879145c7407212ac60fdf84f0bfb43bce52165a75af", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/score.json", "bytes": 5479, "sha256": "10c686674ba8a2727f61d2bc56f5293549f7a561cd1fd8e18596e14989c4a147", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/metrics.json", "bytes": 2375, "sha256": "062a08bb077f99c1b90b61a2473bed409e57179152357cac86a9a075a5236f50", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/facts.json", "bytes": 16495, "sha256": "3c94683808b22119b2faf87681db7611b4fde3a6d26d37f7e34d5d965695e4fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/codex-inline-judge.json", "bytes": 6560, "sha256": "2b8821dde3cbaf2aca077a40b8f99de914b3af4ab4296060f22181c0a1c58713", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-21T17-40-18-776Z-ldhe/postmortem.md", "bytes": 10317, "sha256": "80233a172af42b6f09220d2eb6a507823cf75b23a798e61dc22a0953a5034349", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/ab3f18083d.json", "bytes": 5954, "sha256": "8b6c7907e2e4f41307b8d8c7810d6966616d17b538975a33a530ec3f13528de3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3779, "sha256": "e0bace64577ffc67bfcec67bfcc48b4183cb3bdbd6f5801edeacd29808e7a38c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3779, "sha256": "e0bace64577ffc67bfcec67bfcc48b4183cb3bdbd6f5801edeacd29808e7a38c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/result.json", "bytes": 1462, "sha256": "04737c1bac09a189737a31cca9a65c09e6295f8c6d599a20f0239a40f4ecd3b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/score.json", "bytes": 5225, "sha256": "58c3432023a193db12cae6eeafaefcc06c7280d264ac88a48f913d7af916aff9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/metrics.json", "bytes": 2381, "sha256": "38fd4e23f438497f7abf6633b16e2f3dd643a03e5bffc192546f69500212156a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/facts.json", "bytes": 15226, "sha256": "f6d0a81fbc1fecfe4c3c3559cbf407314be0c76220b1c4271668fe92983e9257", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/codex-inline-judge.json", "bytes": 6635, "sha256": "49e25c5889c78312f7cb26fa222754f2e1db5e0aa5f4f70aa5a316864c78b0e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r2-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-21T17-43-39-100Z-nc1q/postmortem.md", "bytes": 10347, "sha256": "3776311cf891b843cd1ce62ed57267af528b000a914ff22061719562f04474f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/ae6e00bb64.json", "bytes": 6020, "sha256": "b2550ffa28dd612ac3c84aa6bc6fe58071a88ee4d02b19ddd8f183353da17214", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/workspace/tic-tac-toe-game/index.html", "bytes": 3271, "sha256": "233a90d82d52b6db02c95edf890b55c940148c6fe474edf00af9db24144bc032", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/92b0724d3a-tools.jsonl", "bytes": 6917, "sha256": "d736379e1c64e67f1c70ceaa39cde3db78920c9aa924b7053156a364c120bcb2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/result.json", "bytes": 1261, "sha256": "fe928340f20bad6c5fa0a47cef346609f215674e0cc662dcfd09cb392f073366", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/score.json", "bytes": 4725, "sha256": "16b58b59f0d0ab402346faaf1eeca727ec9b36e851a8ad548ad255183d7ac4d8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/metrics.json", "bytes": 2370, "sha256": "91937b1e9fa3e9b22456d4d9f6a71934058d394da7ee5ac027cf917874268d84", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/facts.json", "bytes": 14009, "sha256": "6880ccdc9977fbcf5d3d35936f6805e6c87ef2ece76a50bdee85912675d21423", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/codex-inline-judge.json", "bytes": 3512, "sha256": "880cc8b49030e5607257eadb98f695562e7a4782d68b0ffbd5023486a48da180", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-9b-q4-2026-10-21T17-51-35-006Z-bryf/postmortem.md", "bytes": 10912, "sha256": "dcba26ab2fc27d19c3765aabb2997db9beb8a7d76162280581508df75a44897e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/92b0724d3a.json", "bytes": 4827, "sha256": "45da508de60bb874bd680c6134d1f5aca31121067f20246763a6f41ab373f583", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/92b0724d3a-desktop.png", "bytes": 10075, "sha256": "53dcef7ee7434e19c5665d382dd76ee143416e8b3ee7050e4163f43fb1dffd6e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/92b0724d3a-winner.png", "bytes": 10083, "sha256": "fa9f726c25c01ec50409ed1b4094e232114c330334ac893b54c461390f8d8ef9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/92b0724d3a-mobile.png", "bytes": 7675, "sha256": "75df4f2ad8bd8cb335de446b462a5886b3c4e36889859aa823b5c8f76b978c34", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/workspace/this-work-e/index.html", "bytes": 5147, "sha256": "279a2b5f887bdd8fa3abfb8022e0d5f171123d2db5a79f0f150c583381796b70", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/result.json", "bytes": 866, "sha256": "ca42996dbd22cafe4141e12f93bc8578cfa687e3d5d1595fecfb8c5ebe484cea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/score.json", "bytes": 3090, "sha256": "e3ca5030c5e729539bc3a637a5265656d4aff6b64e08f2605fc0286d2205ac06", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/metrics.json", "bytes": 2374, "sha256": "dab59c331c3f31064efc42c683ceb5e3ae60c6dbccb0122deb53e2bf3230b630", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/facts.json", "bytes": 12539, "sha256": "99523ab77a0fafc67d8e409e419b029572be98cc7b195058c924ba9fc0aee088", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/codex-inline-judge.json", "bytes": 3103, "sha256": "172e978025799d9bb0165254143311d3b7aecd0f2c049a6958a2aca10543ca89", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tictactoe/tictactoe/tictactoe-ornith1-5-9b-q4-2026-10-21T17-56-17-279Z-61zg/postmortem.md", "bytes": 8674, "sha256": "b6184acfd71d63084af5c9264a0dd69f689356302e02070e7f08d253478763e9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8c44bfb429.json", "bytes": 4098, "sha256": "a947da68ae6a9e0db122262aef4c4a6f001ecc4decb07d3bad2a6b64d4d31117", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8c44bfb429-desktop.png", "bytes": 11727, "sha256": "becc56beb86b059a94cca3efe11526ed9a3b52f724ee99dca9f8c68a2a0d19a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8c44bfb429-winner.png", "bytes": 16562, "sha256": "b0c8efec848ebf84c7f30d94dd81a1dcbf2335010bded6ae4acafa4e19f30c17", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/8c44bfb429-mobile.png", "bytes": 9403, "sha256": "c6ced07ed4b04dcaee2fccc16369d143a8f4ec37ed1c5c1b2e64b33ace78cc04", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/workspace/tic-tac-toe-game/index.html", "bytes": 4657, "sha256": "0d04645c36ff52cebcc89005ab2cc37fe610e5c6f9807698e2160036f5b0102f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/result.json", "bytes": 859, "sha256": "c4d4e0dcdf95ed989d161308a97fbbe8d75d6e07e78481c2851288cae2ff3401", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/score.json", "bytes": 3088, "sha256": "c2d1fa4d4a76f3cce12e3a010d80b82cd9920e6346b5df3fbc780e73515b8355", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/metrics.json", "bytes": 2365, "sha256": "bfedb628fee717d2db8c436a24154ea4d5db83f0a4a12a1c4d86c7995366bbb6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/facts.json", "bytes": 11724, "sha256": "36174b8b5e2b2ba07c86723a2d406a7f88da6be9d3798eaf65d9eb08572c8270", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/codex-inline-judge.json", "bytes": 3094, "sha256": "fd6a4ae78f146cf19b60b4dc0d27c9049e8b80d48d6a04db6989b0203f5bf047", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tictactoe/tictactoe/tictactoe-qwen3-5-4b-q4-2026-10-21T18-00-38-024Z-n0jh/postmortem.md", "bytes": 8651, "sha256": "14ec397fffff110561b7ae6ff0332e70544e81e1071b2b01d67ce89556b76393", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d13578715e.json", "bytes": 4009, "sha256": "1833edac3159af7b2d5e7910571dfdfe7947ea3969e37b918efb50fc6a20ab13", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d13578715e-desktop.png", "bytes": 12239, "sha256": "56a9f1274ce610395808788e58539d6461279498d7373c05e6cc6e511ae47ef5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d13578715e-winner.png", "bytes": 16082, "sha256": "efdaec507b6bd6b4debd69742eb165dfceac6d42d3c42d59e3561ee3088e08be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/d13578715e-mobile.png", "bytes": 9777, "sha256": "d636aeeb5afe4e7e63a8c4946bd767bf510f8358edf7651fc0eb6c41a3e4e204", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/workspace/tic-tac-toe-game/index.html", "bytes": 3970, "sha256": "9a6032e992b37389629a2b120b9b034ec09fa15826a8845d15b2ca4287b54852", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/result.json", "bytes": 932, "sha256": "9b44d713d53bad0b8a9e08527b21e50e6845f85056df8be1536d6ba033c84b87", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/score.json", "bytes": 3327, "sha256": "63a87d821c7dafb975384850beb550b4d41ed74779f6c9785461ef8a3ea42fee", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/metrics.json", "bytes": 2399, "sha256": "b603bbeba66c2e68597c14550c32822aa45860f4cf71d62ef4042cc5ac518e54", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/facts.json", "bytes": 12119, "sha256": "cfd0b2d0006ef6686788e813da540e3612509a2556a4f7f686edeba1e0b01747", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/codex-inline-judge.json", "bytes": 3216, "sha256": "9ceda6f9c2b33adcd34f6ec97b0d9f557f71f641b28252e2e840843868676f57", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tictactoe/tictactoe/tictactoe-underdog-saluki-27b-iq2-mix-2026-10-21T18-02-38-810Z-5y5i/postmortem.md", "bytes": 8912, "sha256": "c58f54769e2201e1dae9a172b3cddbe8417de37e3ea8aec9fdf728a93fc4265a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ebc2863434.json", "bytes": 4232, "sha256": "85479e263bc5a76eb99fd3471a58d9c87a1418db6ef83bae02fcfc73cc2ae627", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ebc2863434-desktop.png", "bytes": 13224, "sha256": "ef7f1c8f86f84e5dfd71fa9be99aab1975f874da85255b46a3975e0996090df1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ebc2863434-winner.png", "bytes": 16529, "sha256": "2975d03da9cce2b56722b873e53c0beffe7c64bc98e9973e093b1d84169dc7bb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ebc2863434-mobile.png", "bytes": 10479, "sha256": "4111065ad6a76ec9ac64a0c724362b5912200ad2b392ce131598a6dde04259ac", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/workspace/tic-tac-toe-game/index.html", "bytes": 2741, "sha256": "ea34bc102da3388ee71161335f4b30d675de61509c8019623b93b9a7dccf1e84", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/result.json", "bytes": 867, "sha256": "18e27fedf7e57f5bf5764ee8552dc1a36738353e0ff0dcdc38c26c514bf95606", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/score.json", "bytes": 3302, "sha256": "656cbd39f078ea3c2ba3c929e4a0f47215b7cec43db482f26bdc01bf847c7631", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/metrics.json", "bytes": 2364, "sha256": "195e1051bea49220732e568df4dbdb23e6f2b6701c9333ca5feb961a683eb02b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/facts.json", "bytes": 11705, "sha256": "c458e8bc0b9f63ab48d9483788cbb0030a708208af4723076382f3a6ad58cb96", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/codex-inline-judge.json", "bytes": 3104, "sha256": "e338e24984ab2d5ba575e77f5d92c5f9bbb1af85acff7f3aebba64ca2203f1f6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tictactoe/tictactoe/tictactoe-qwen3-8-27b-q2-2026-10-21T18-07-27-073Z-ovbm/postmortem.md", "bytes": 8877, "sha256": "c953acbe0db1fe88eb4266b34a9fcf16ad8e1c1fe5c4bd8fa38683eca8e988cf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1aad1070d3.json", "bytes": 3993, "sha256": "98669da35a8a83e28ce5a32f3ca400647ea472c3f51dca6ee417f768b91ecf74", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1aad1070d3-desktop.png", "bytes": 11736, "sha256": "39fbe0a853d30db10252e061a6d82064f798b557feecb4a1a8643540764e2c6b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1aad1070d3-winner.png", "bytes": 13569, "sha256": "9b6537747befdbd3b6bec7cab62ec9770224bb236950f3eadecd58977e8b9cbf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1aad1070d3-mobile.png", "bytes": 9546, "sha256": "991e79dac1836b1d76dc92bc4909c0f64e5c394797189fb033f5813f7d660527", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/workspace/tic-tac-toe-game/index.html", "bytes": 4776, "sha256": "01e10079a727be2cedbe58d15e7378b6b5950316390c53f53a70717d4b7d5c6c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/result.json", "bytes": 861, "sha256": "90037b8f98dd4b9668aab5b8505f602da3a5f8ae815ab53f4fee769ce978ccc0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/score.json", "bytes": 3088, "sha256": "0fe08cc49656260bf0e3050c2912fd3a3545ef87159d811656d3cc1b5bffd3fb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/metrics.json", "bytes": 2370, "sha256": "cdcf7adac0579040139afd55c8dd2d8095991d905de92e218b0da9dfa79587a7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/facts.json", "bytes": 11932, "sha256": "5a3d2432e404dbece0a043772a66b1a10b97d2ee9fc1873629f4169b29a0199f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/codex-inline-judge.json", "bytes": 3089, "sha256": "b37302790ad144797aa0806290c377694c3348464cb9a20bb0cabdb59f08660d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-12b-q4-2026-10-21T18-16-18-511Z-7dxe/postmortem.md", "bytes": 8641, "sha256": "520d832e1a81206cfecb5a94253109039cba5918818830131fca9864a14bf773", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4a7a20716f.json", "bytes": 4119, "sha256": "aadefb98c2f79df6718aa810d6cd15847b9660ad9bd0b0f878af16bed5fbb039", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4a7a20716f-desktop.png", "bytes": 12052, "sha256": "2425bcb224318ee1e7e5eddca7c06e9b454a476029f018751bb38226432e709d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4a7a20716f-winner.png", "bytes": 15201, "sha256": "82de4cf360a0ae43cbc5ca1583b1475205bcc6776d5a5b9a5aecf00894e5c42b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4a7a20716f-mobile.png", "bytes": 9588, "sha256": "53738218c22466d86a15ef76ea92648157dd7b020a0b560b535ef15920b385a2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/workspace/tic-tac-toe-game/index.html", "bytes": 4000, "sha256": "7a993fdf75b226c7cd8ed48751fd469705a924bf5dcae5e6e8aa1b5a28449dce", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/result.json", "bytes": 860, "sha256": "3a2d97c3ac86957c4d26e6c96341dde9442175b0a9b243b8376c5f87f9ad163a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/score.json", "bytes": 3297, "sha256": "9c6148b2f6161d6ed03876094370a817498f2084fc3bdf023dba4dee0841f4e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/metrics.json", "bytes": 2334, "sha256": "7cc36928b99c1710a8ae6954fc084eab5190a6512900d4a93b49c492215bdc50", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/facts.json", "bytes": 13311, "sha256": "cb1253c06a31e48897dac6c0f260ca258e5a7871b36129eb2c8bfad97157e2f6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/codex-inline-judge.json", "bytes": 3096, "sha256": "50a0563a7c5a8de1258664429a98b4bd4b67dd14c128cfa1e6c469a48d2439e9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tictactoe/tictactoe/tictactoe-gemma4-e4b-q4-2026-10-21T18-18-36-731Z-xnr7/postmortem.md", "bytes": 8872, "sha256": "fe8a7a64c046480a489dc5c847a8493b9971e5a0dc65ee031aade1641ec5268d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/c61eb7689b.json", "bytes": 3990, "sha256": "c335a7386fb0394cf97c091a930c8f3ee419c6a29f52145dda55a422587b250b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/c61eb7689b-desktop.png", "bytes": 7299, "sha256": "b5329d6a597249c2db9404d329b5c6cd5c150259ac89105f4de700b636be64df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/c61eb7689b-winner.png", "bytes": 9719, "sha256": "7c687e3d68c583521b0919b648b1bed64f48ba890550bea1a5634690846605a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/c61eb7689b-mobile.png", "bytes": 4951, "sha256": "2d4d4c31eba182aa41803d5e1b11ff06050496ea1a4074c5089989782bd81fad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/workspace/build-me-a-single-page-pet-shop-website/index.html", "bytes": 2078, "sha256": "8d43af7138c0b26c2aedc5bf56fe7661f65983ab8a5b337d1d345c9c41634035", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782-tools.jsonl", "bytes": 5220, "sha256": "51289b2d23ef732bf4b3d7298c08305da12da4e32e18e03f03a20691467ca564", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/workspace/build-me-a-single-page-pet-shop-website/index.html", "bytes": 2078, "sha256": "8d43af7138c0b26c2aedc5bf56fe7661f65983ab8a5b337d1d345c9c41634035", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/result.json", "bytes": 899, "sha256": "bcb6ffe470b6ace2de551a9568d15a227745eff3e84e8301dea5b8b22200a9ed", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/score.json", "bytes": 3682, "sha256": "69b81961da83a2d93cd00ec517d60a39bd4625eb5debf0f05eb63a5918f69473", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/metrics.json", "bytes": 2387, "sha256": "da1d61a9c4b065144afc2ea8786680c45149d3ec66b7699b6b459ece8aed273c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/facts.json", "bytes": 13964, "sha256": "f0f755ccb272a04e2400c492ef9d8984ebb93ad34dd0eb7ce2d700cd2a90e75b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/codex-inline-judge.json", "bytes": 3719, "sha256": "4340bc75e0a34c862f3bf5d9632b98722dff069756c60c959da754e97cc16616", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-petshop/petshop/petshop-ornith1-5-9b-q4-2026-10-21T18-21-24-432Z-kcbd/postmortem.md", "bytes": 9197, "sha256": "60f15b51f39e3234a0242fb4f96133dae23c3c4e4f69a684fc31ed4b76d08e8a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782.json", "bytes": 3995, "sha256": "1b223beaf7852fefcba003b3a139c1363892084707a559a2e21be6bc901227b8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782-desktop.png", "bytes": 1645183, "sha256": "892171b6a13acd13238a2b771091c28e5e03270310920e97b3a8794a98b4fea1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782-desktop-viewport.png", "bytes": 861367, "sha256": "dd7bb8f9a271637248df84fcc8a991c42f792aa643a3d2755b8064fcc6a9ca38", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782-mobile.png", "bytes": 1647387, "sha256": "5a5bc6f683a22abff78cf1af040a31837691cebfa5707d5db2abad6530edc716", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4afc35d782-mobile-viewport.png", "bytes": 173286, "sha256": "424b1ac5f38dfb978fc034dba01ede1e11c721c82fd87d945d018997db2d7ef5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/workspace/pet-shop-website/index.html", "bytes": 8680, "sha256": "a5f73d1b1f5c80badaaf2b491c3771005563e9b27690843730fe420acaca41da", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/workspace/pet-shop-website/assets/logo.svg", "bytes": 1360, "sha256": "448e6e2663f6d4a0b046cfe11a9fab26b7b245a26ff36ca345142947abb264de", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/178f97af19-tools.jsonl", "bytes": 5846, "sha256": "492d38f3d7897f9b201ca5936ab63ed373c7ca2badc5535837a3ae466f801e5c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/workspace/pet-shop-website/index.html", "bytes": 8680, "sha256": "a5f73d1b1f5c80badaaf2b491c3771005563e9b27690843730fe420acaca41da", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/workspace/pet-shop-website/assets/logo.svg", "bytes": 1360, "sha256": "448e6e2663f6d4a0b046cfe11a9fab26b7b245a26ff36ca345142947abb264de", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/result.json", "bytes": 1088, "sha256": "f55c7b811eb804fb0a50a45ac1131a424a1107465e1d8f958f3d6e926451fdf1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/score.json", "bytes": 4420, "sha256": "ead24cd6cb8d05d43c26827486e3c353064265175dfd2f230cf8f0e33f0cdcc9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/metrics.json", "bytes": 2388, "sha256": "fe90dae21ecdc9e134c80bb54438cbab0c9038de24b32022505bdf1e194708db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/facts.json", "bytes": 14226, "sha256": "9ad053c49b1e098e4c40d236d0abc2a66e7753a431bf1abb4bca6c9d6e9b775c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/codex-inline-judge.json", "bytes": 4030, "sha256": "135c1e9bef3bdafcc0c8f8cb91e5a85002ed491e6ff4ef9d9ee76bcb2accd74d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-petshop/petshop/petshop-qwen3-5-4b-q4-2026-10-21T18-30-42-053Z-8lcz/postmortem.md", "bytes": 10792, "sha256": "ba5bc3ec709c3aef7d6dfbe4596394b12de801e631688759381aaffbb919a809", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/178f97af19.json", "bytes": 4567, "sha256": "c5a4929e11ef9deea7db85ea5e2acc73972addf04409c36f08b69013231e4324", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/178f97af19-desktop.png", "bytes": 57389, "sha256": "05b597a12bd7ad2dec788d35fe7fd5d5de9fc2a9a51ba628b91af437dbc8fbb4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/178f97af19-mobile.png", "bytes": 59884, "sha256": "d03659f69d7b0982576289524c51d93789ea3203550f98fde6bf5a1a1feb1572", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/workspace/pet-shop-website/index.html", "bytes": 6894, "sha256": "8f43c2c4f3c4c94695abb6162f8543068ba2ed930b9ea9dabba125a3f9665942", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/69fc143957-tools.jsonl", "bytes": 4829, "sha256": "7a618af6642cccdc9a4f73f7a89155e9542ff48b1f49dd8787be2324aa67b73e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/workspace/pet-shop-website/index.html", "bytes": 6894, "sha256": "8f43c2c4f3c4c94695abb6162f8543068ba2ed930b9ea9dabba125a3f9665942", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/result.json", "bytes": 1221, "sha256": "42904df9927b1a021c71fa2df14c12ff4b0de1b51cc0a389e9289e448e44eaf8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/score.json", "bytes": 4454, "sha256": "94b13d76b3dff4e2bfd20bbfca2f29531dd86ddef4acaab2cf2235f6d525962d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/metrics.json", "bytes": 2414, "sha256": "042d8b3e59306885474f9d5d31233b9165ef1798681c144790f0b67a71ca2d23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/facts.json", "bytes": 13839, "sha256": "0dd570f9278229262feeeed4cd21482b9acb0d39b894e2db88d7332c6e085719", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/codex-inline-judge.json", "bytes": 4021, "sha256": "78f79b51855877441f50117efbb01478c147a8b8be4cfe32ea8a148e507261cd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-petshop/petshop/petshop-underdog-saluki-27b-iq2-mix-2026-10-21T18-35-19-913Z-26sg/postmortem.md", "bytes": 10831, "sha256": "40ad3ccf3637fb66c59a0b84cf1222ba01cda1a63c4f572398411f136d183c6d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/69fc143957.json", "bytes": 6493, "sha256": "3066f6e9858599cf27824cdd70703b36b312a3a09dc5822f24bce6bf0c9f66a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/69fc143957-desktop.png", "bytes": 97283, "sha256": "1c5a08d20956d9eed53e0c15305a4ccb50b7c5d987fc73d3a4a8ef3515b53246", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/69fc143957-mobile.png", "bytes": 89414, "sha256": "0fd979d72ad6da5e52205b6fad04b66e4cf94be5e8fd741484124543171eb859", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/69fc143957-mobile-viewport.png", "bytes": 33023, "sha256": "3da24310ddd81adbabfc97c9f7be35ef5c587288a16053fe1aa9c5bd1bc267a6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/workspace/pet-shop-website/index.html", "bytes": 6137, "sha256": "a32fb188684b598c19309449778c5b3ee78ec68962bb36a3d6a45da2a70cdc7b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed737b6d12-tools.jsonl", "bytes": 1639, "sha256": "3d1f99dfd418a9b28b2647351835d168e3ad27fe1dc5f0510abbd208f278d863", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/workspace/pet-shop-website/index.html", "bytes": 6137, "sha256": "a32fb188684b598c19309449778c5b3ee78ec68962bb36a3d6a45da2a70cdc7b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/result.json", "bytes": 1096, "sha256": "a5450195ce6a28b7b8fd003ee7eb8b188e302a5e7157a54a9642b298430c3a87", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/score.json", "bytes": 4317, "sha256": "a7731d7141378cf1c662122a475cf282fb77ce7ae11e0dd03358d9ad02c45718", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/metrics.json", "bytes": 2378, "sha256": "aac3def9b7e330b45c993ed6e7a87cefb571bc76a78be8cabcdd3bc2771d409b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/facts.json", "bytes": 13756, "sha256": "30c7d756fd38bd81459b17b80ba2f8a7b36ddc2b37f148bcf78a8ed7afd3b2c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/codex-inline-judge.json", "bytes": 3293, "sha256": "39922bdb47ff621d0b42eba5bb900601ab4ef593ae99802823ff5a35a467a5ee", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-petshop/petshop/petshop-qwen3-8-27b-q2-2026-10-21T19-24-09-759Z-xb5z/postmortem.md", "bytes": 10510, "sha256": "28d405361da787a18ff6a9c85437e626cd784cd33c9dc434b0677900c13e5d8f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed737b6d12.json", "bytes": 4063, "sha256": "ab54df4de147c642ec5ad66da58316f6efaccff73fe79e427e55ae975c6c0e1f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed737b6d12-desktop.png", "bytes": 135100, "sha256": "957a6b8b112453b7f9a933476d9616c9910f8b3c55f38793fcaec4dd4ab57c1e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed737b6d12-mobile.png", "bytes": 144790, "sha256": "55d38dd6c26950735d29619137d9a3a020fd79202b7ecc33c2268d8281762785", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/workspace/pet-shop-website/index.html", "bytes": 8900, "sha256": "935eec522492b7a39cc99739e9eae829f6fca0ea32d5556572acc70f3024cade", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/96caa9b603-tools.jsonl", "bytes": 10036, "sha256": "94dd8a5b544ad196c4ce85c300f8da56682908623d6f3fa729dce150ee5ac3b2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/workspace/pet-shop-website/index.html", "bytes": 8900, "sha256": "935eec522492b7a39cc99739e9eae829f6fca0ea32d5556572acc70f3024cade", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/result.json", "bytes": 1090, "sha256": "21766b48d101b2031223497ba4debb0bac5f58646072d2227d9d94d5931a33a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/score.json", "bytes": 4284, "sha256": "156b9387605f3287656ede7f09b2a8aa2e5326d7d0cb4892b901e97ce7a9642e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/metrics.json", "bytes": 2388, "sha256": "7789136f9b49a29b9f19396ca8a1b7355dda5dfc99d2fde5185f5efa0be0d2ab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/facts.json", "bytes": 14586, "sha256": "b842bfa36888380d585184efb87e369c009c0abc682f5e5f9aed1031df951933", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/codex-inline-judge.json", "bytes": 3373, "sha256": "f8e12e7c77cbcd504da9ae8d3e62d11d05029bf0862e477f76fd3f7d30f98a0f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-petshop/petshop/petshop-gemma4-12b-q4-2026-10-21T20-50-24-064Z-3mp8/postmortem.md", "bytes": 10481, "sha256": "89bc4c0437d1451a3662e80fa4c0fd482637483ee94b74b28206b55d9101c2e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/96caa9b603.json", "bytes": 3980, "sha256": "083621fc96f9fe737d4acfa8d842897ff69f656820d0b557b926a754a9ae5b10", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/96caa9b603-desktop.png", "bytes": 94166, "sha256": "a83f1d672a108532044d23bb5ff37e5ac06a11fd534a166ab0ea4c3d75254ee2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/96caa9b603-mobile.png", "bytes": 87970, "sha256": "af66780bb8ee194b73b4df299c50aa54511345510f017703b809b90d0a73fbc1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/workspace/pet-shop-website-build/index.html", "bytes": 4570, "sha256": "7396c5daa4b173b35d4bf9879ba5a3e470846c602d9d3cba4b6399aad57efdbc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/workspace/pet-shop-website-build/index.html", "bytes": 4570, "sha256": "7396c5daa4b173b35d4bf9879ba5a3e470846c602d9d3cba4b6399aad57efdbc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/result.json", "bytes": 872, "sha256": "fefc32d863a19de0f51cf23c33c103112d3e87a009253a1d9cd3eec1046c1c40", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/score.json", "bytes": 3594, "sha256": "2d853642b7c2291dee38a476a4498ee32d7ae7341dad8dd024f143cde5c0ff99", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/metrics.json", "bytes": 2377, "sha256": "3cdd9e9bde9f1f07fefdf14c34b1cb7beb4f0084e7c4e1a13324fd64d1ac609c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/facts.json", "bytes": 14320, "sha256": "e0537218ce74d0eede0e084e426e7da853c15fa5f50b844f2287f240d60eb6f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/codex-inline-judge.json", "bytes": 3032, "sha256": "b05dfe3c9e361a09357524e8e022d0569c7f025863ecd7faeccc3bdbdce29d72", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-petshop/petshop/petshop-gemma4-e4b-q4-2026-10-21T21-00-41-035Z-spos/postmortem.md", "bytes": 9095, "sha256": "c1285afd1c6208ea8a3941943ff7baaa58a785083a0294489d4b711caade15e6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/824365ad63.json", "bytes": 3814, "sha256": "91988c2a9c224db0b78c9b82d1a49c7e4a61fa5f68142600a325e96bbd550f66", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/824365ad63-desktop.png", "bytes": 87784, "sha256": "6a7b24be38d1263553c50a824ff14e7eabb3526ef17acc2d6bf8fd999b668f25", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/824365ad63-mobile.png", "bytes": 92447, "sha256": "837ebef0c7fa76de0cd18056be375c140ebd38ce3f8256f401bae24174c25847", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/workspace/pet-shop-website-with-logo/index.html", "bytes": 5322, "sha256": "92a0d66178edd0d9994350a1a32b881a2d54728ecfaf76ca32229c424616aa4d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/workspace/pet-shop-website-with-logo/index.html", "bytes": 5322, "sha256": "92a0d66178edd0d9994350a1a32b881a2d54728ecfaf76ca32229c424616aa4d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/result.json", "bytes": 876, "sha256": "f76b8dc30fa66f4fff832d920e92050e3ed9d3d8edbb6984f261bb78b6b81a4c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/score.json", "bytes": 3625, "sha256": "d6cbaee8f4cde4b10f4ea7ade488253b22f6677d81e660ecb1dae3831374e650", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/metrics.json", "bytes": 2379, "sha256": "3d62859da26443af12f7fbba69ad7f40facdb880381e059f1e2ba190a23ef341", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/facts.json", "bytes": 14949, "sha256": "1a9c1d20d96812decd5c6d33aedb75291d91ccfd570d02e99debf1224ec8f3e7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/codex-inline-judge.json", "bytes": 3050, "sha256": "978fac4bc2315b09f4baea265c42fdc6958219c33036d8cbfdf18e6956c19d64", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-petshop/petshop/petshop-qwen3-5-9b-q4-2026-10-21T21-04-46-650Z-q0kx/postmortem.md", "bytes": 9224, "sha256": "0c5455be9fb1d2173bb04e5705081e969fba680c8397c56b38a6f05074ba8e20", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/77ace886ab.json", "bytes": 3947, "sha256": "dbecbd426d877360ac8051a3e8047b99eb1a9a2e94559f597ad3817ed5eb45ad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/77ace886ab-desktop.png", "bytes": 84230, "sha256": "00987913d00c3777bb6fbac6fea4d3e5041a66c29c0845e82c12785c7494802c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/77ace886ab-mobile.png", "bytes": 84331, "sha256": "c6b198c4caf59f336a6b762dae7d46cd2cd2da1458b9a53a1d45c18eb303d7db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/workspace/tank-combat-arcade-4/index.html", "bytes": 8017, "sha256": "60be734f2fd162558709653811b9157559438c39e1801ba57e4305a94f6aaece", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/workspace/tank-combat-arcade-4/index.html", "bytes": 8017, "sha256": "60be734f2fd162558709653811b9157559438c39e1801ba57e4305a94f6aaece", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/result.json", "bytes": 925, "sha256": "fd53aed33f5ec2e08cb1c59202ec513bff0cc5ee60f8a3534568744cdc8743b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/score.json", "bytes": 3213, "sha256": "b38110fbc1c4498be79e3a7b7180941f1e7d9232c6f4a20197da7d95d313fc7d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/metrics.json", "bytes": 2343, "sha256": "497f89c3c1862d58e2e5003dabad7c37a173e7c0e5b90baa26839a82f195dce3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/facts.json", "bytes": 16069, "sha256": "0ebc6a9956904c8cdecf5094cc3cd57805f6fb073bb951a109406eaf9a20c6f2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/codex-inline-judge.json", "bytes": 3376, "sha256": "6ad20b371f96c07c65d7a369bae2aac510edb658596d6541f651704e84ae6866", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-4b-q4-2026-10-21T21-11-09-303Z-bcf4/postmortem.md", "bytes": 8737, "sha256": "10ab6e0b3f38f55758e30ca348206d99c79b7b66de9747466b5e1dd0a3a5d65c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1da550bc0b.json", "bytes": 12065, "sha256": "da9952fad956ce7cc3250080bf202a270a8114df9f859a6d987fcc47a87b298e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1da550bc0b-desktop.png", "bytes": 13732, "sha256": "8a281943143b255e9a4dbabb742ccd4d6fc3ab7a88416e3bd4deab1c01f9c24d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1da550bc0b-after-input.png", "bytes": 13762, "sha256": "f3c759ade53c007153898dca49fdfaf1aebdfc2fa0e8f3ad4732ac335941a824", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1da550bc0b-natural-play.png", "bytes": 13731, "sha256": "92e03fce180b9938bc682f63868b88d82b08e544720fb1c32a4b339ca81f77ad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/1da550bc0b-mobile.png", "bytes": 8104, "sha256": "32b71a4af822b6fd626d272b83b08dac319cec411054c89065cec1cd04ae29ad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/workspace/tank-combat-arcade/index.html", "bytes": 8583, "sha256": "f4ac8295e01f165364f78ca102c459acaa8bc5ea7253ba24a3bc77cc5ff56852", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/workspace/tank-combat-arcade/index.html", "bytes": 8583, "sha256": "f4ac8295e01f165364f78ca102c459acaa8bc5ea7253ba24a3bc77cc5ff56852", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/result.json", "bytes": 995, "sha256": "fa6af4d2eae5c474616640769cdaa70b33630a0fb7dfec11f46f06c514aca3eb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/score.json", "bytes": 3226, "sha256": "54c09205fc1e1cc1a440cc2a79fdd5770f1491a1502e83e51407d4879e8c546f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/metrics.json", "bytes": 2403, "sha256": "023b24308946f577b4a6ce97cc8b0ddcfdb9f00f5337de05ddb7f9f1031e5cfd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/facts.json", "bytes": 12701, "sha256": "e97e624274ccdab5caf78229c83246e7dcd34f690503ed11676604ec300b1839", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/codex-inline-judge.json", "bytes": 3485, "sha256": "ccdb4428778383496f9bdb80ab84d52c89da520ad2d6db197cd67b8bd1aa8649", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-tankcombat/tankcombat/tankcombat-underdog-saluki-27b-iq2-mix-2026-10-21T21-24-38-138Z-69c0/postmortem.md", "bytes": 8688, "sha256": "64eea675b7f98fa5effd0216f0e1d3da894a480cbac3726952e93ea84ce42e4d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4f43c38a23.json", "bytes": 16305, "sha256": "b144015a4ca42faab3b62a18cc126c592b60ad882f964086634c1ed0be315f45", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4f43c38a23-desktop.png", "bytes": 20656, "sha256": "e511d2c7ac3afc111fc2697cf6fcd96571115aed4e5e597b4067eb2267da0e74", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4f43c38a23-after-input.png", "bytes": 20634, "sha256": "23ce98d2121a157c43359c4c3942d712e6928bd52f58ad53044ce7f5297cc834", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4f43c38a23-natural-play.png", "bytes": 20528, "sha256": "84c8b7ce8333c24f5a8c1ac970fc7984bcc4645d1053d76f1bdfa51cd5bfd4f3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/4f43c38a23-mobile.png", "bytes": 13502, "sha256": "188b47362e79c4737e1982c96b36aed0577311477726c81daa6bde4ad1cbd812", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/workspace/tank-combat-arcade/index.html", "bytes": 7235, "sha256": "3571c32cd5be9f459435b3a4926383b4e2ff487cb29bfe4bde267cfe72b954b9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/workspace/tank-combat-arcade/index.html", "bytes": 7235, "sha256": "3571c32cd5be9f459435b3a4926383b4e2ff487cb29bfe4bde267cfe72b954b9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/result.json", "bytes": 931, "sha256": "49e9d0ad9a4431aaf17ced4ac03f9ad0bdb4af4cd53ce23fac12f7a505cae772", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/score.json", "bytes": 3203, "sha256": "dd33a30b3fb544bcf32aac7b5d71455aab0ff76f5107422270ed095b9a919074", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/metrics.json", "bytes": 2376, "sha256": "3e078ebe70806cc7281c29cad1e361f1f66ae678d4746a368d165e490478cde4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/facts.json", "bytes": 12555, "sha256": "7483870ed810a28ab7404a90cd8804650457b668df1b7f7ff57be8c5f1ce299e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/codex-inline-judge.json", "bytes": 3260, "sha256": "b7698bf9a924edad11b29160c270e552b27660df0fb41b6438e251ab0645c0c9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-tankcombat/tankcombat/tankcombat-qwen3-8-27b-q2-2026-10-21T21-34-19-384Z-nzrr/postmortem.md", "bytes": 8668, "sha256": "229e8decc0b158f7796ecd3ad862afe584fe8a845c4d9f95373276e1cd40fca3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/a2ee0048c9.json", "bytes": 6379, "sha256": "d668a599e163970def3e20a3352a560e5e1bc3e47f32ae571556fd26630e8267", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/a2ee0048c9-desktop.png", "bytes": 12287, "sha256": "1a7f6ef95b23adb6795e745741e2db29466b20f3e684fe7443852a63e0f80b2d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/a2ee0048c9-after-input.png", "bytes": 12225, "sha256": "1b3e8f2850461745cadb647024cc3bb6f97600451e8252197095f79b8463c6a1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/a2ee0048c9-natural-play.png", "bytes": 17671, "sha256": "f11bf679ebcaba5031b286b98bd2f872c1d41e72bf6dc193c48a1cd1c267776a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/a2ee0048c9-mobile.png", "bytes": 9174, "sha256": "0d6149cae8231a13b06570fe7e2c5904ed6c98cb43c84bbffec1b581f4e4923a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/workspace/tank-combat-arcade/index.html", "bytes": 7616, "sha256": "e1977d98f0040fd3fc5cb3bbdede8d74292db14f201b5709ec1f875dca0c9538", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/workspace/tank-combat-arcade/index.html", "bytes": 7616, "sha256": "e1977d98f0040fd3fc5cb3bbdede8d74292db14f201b5709ec1f875dca0c9538", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/result.json", "bytes": 925, "sha256": "27b99ae2fa3df738b1ace45918bf6e4de7062f9d2fc73172c625373b88175423", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/score.json", "bytes": 3196, "sha256": "a8ae7176dced7c81449d03688b63af10164b8e319ab563a77bf5ea53721a22fb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/metrics.json", "bytes": 2339, "sha256": "f9c5ac9427705c1067515bef8ef1a7b03265b27072052a6f81d486816c084bd9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/facts.json", "bytes": 13434, "sha256": "3b148b1f11295bd327e3c07d5536b0d3f11639ed7d1a4cd468b690aebb4218d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/codex-inline-judge.json", "bytes": 3323, "sha256": "a321fe807e9ca1605be583b1cc7f066e4936c1a5dfcc38c1a001f1f8a083b0d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-12b-q4-2026-10-21T21-56-14-914Z-jz3o/postmortem.md", "bytes": 8696, "sha256": "b36a61a1a8f636427c1d1aa6c84aab23e54414d2ed8745aee1e5d4121b6e7c23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2f94943bbb.json", "bytes": 9710, "sha256": "2916dadf5c2aa64673230bed8625763abcd72832155d8c4d2d500c2a06e82b04", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2f94943bbb-desktop.png", "bytes": 12696, "sha256": "47f700c1fe9e74f196d6aed63032af4544486f3e765f97fc93ea7f944b4f52c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2f94943bbb-after-input.png", "bytes": 15158, "sha256": "57e4daaee0f2949213f93576968060151db92bcf0afed058e6ffae556630ca54", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2f94943bbb-natural-play.png", "bytes": 14807, "sha256": "75f8c01dc832bcb2bb6e6bb6c0cc553c0705afd21e56bb9dcb8264f23e3dc762", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2f94943bbb-mobile.png", "bytes": 10218, "sha256": "8d3be057475128e2862e903bafb3efea80ce1e5ffde3cfa8594134d31f0b56ea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/workspace/tank-combat-arcade/index.html", "bytes": 8567, "sha256": "198ba0bacfb3838123da640902b557950951b142b0899a202bb2f9c844518610", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/workspace/tank-combat-arcade/index.html", "bytes": 8567, "sha256": "198ba0bacfb3838123da640902b557950951b142b0899a202bb2f9c844518610", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/result.json", "bytes": 923, "sha256": "b82cc43772c18a47ebc6f447b9dacadadf221ae78dd94aca8156e3a09078fc90", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/score.json", "bytes": 3186, "sha256": "3893b2fdff8ea3517c8feb27ff19ed4a385f4a86453f9f24797579352c7d2b59", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/metrics.json", "bytes": 2372, "sha256": "a3d4220700feff1abef9c223671cceeb63a5437087409caafbf4772e8d7ad46a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/facts.json", "bytes": 12230, "sha256": "542e74e9c87407b2523fe6efd074f770b0f1cb7bf0fdc8bbb09634cd7930f125", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/codex-inline-judge.json", "bytes": 3343, "sha256": "370303bdabfd901c35c10e58788acd4887f3029d7a4a4843a605f4308b1bc055", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-tankcombat/tankcombat/tankcombat-gemma4-e4b-q4-2026-10-21T21-59-53-810Z-mpoh/postmortem.md", "bytes": 8659, "sha256": "06bd7ee813e242a99445359492593053296f9c2059c323b5204e283c2acab03d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2bfa12878f.json", "bytes": 11342, "sha256": "22f97a2118739ee81fdf438a9abfa2288d0e3eeb067db469dd8229119dcd14e5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2bfa12878f-desktop.png", "bytes": 7058, "sha256": "0f459b0bb62d6ac1416e2cfb70787d06ce9bcfa64f769d43f43acbfee6e7ec32", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2bfa12878f-after-input.png", "bytes": 7159, "sha256": "cbb258c7fb4efca54fc05a542ec34f741018700c47e9690eb490f04cca2cfb88", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2bfa12878f-natural-play.png", "bytes": 7031, "sha256": "b44d122a7a1b1fb5f66e08d97bc8552b4e4fcef2a417c8e1140ba1b31001553c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/2bfa12878f-mobile.png", "bytes": 5582, "sha256": "0521a7420fe2351682505c876b1058e30a7e40c4fb21fcc94834898b92581e4f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/workspace/tank-combat-arcade/index.html", "bytes": 5723, "sha256": "eb83bbecefe113cf69ad3ff2d5e6580156f180cffa8008560e4f5c88530d5106", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8-runtime-log.txt", "bytes": 2681, "sha256": "a56c7a5f7df9d50ba282688c9c790b0c2020924a30a52df1e692a20d59ae2657", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/workspace/tank-combat-arcade/index.html", "bytes": 5723, "sha256": "eb83bbecefe113cf69ad3ff2d5e6580156f180cffa8008560e4f5c88530d5106", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/result.json", "bytes": 1197, "sha256": "d9fd2e4395f53a17d9cc7614afe6a83d182fa7932577bece99355376e3f3d84e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/score.json", "bytes": 4677, "sha256": "d9012efb0059f2345c788c00b42350ce59e38e696e6247b50e1aca33c712aaf5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/metrics.json", "bytes": 2345, "sha256": "df1551cf4f619589acbd04147dcb9d38cd3a0a5e9fc186b35e710e26de262e8e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/facts.json", "bytes": 17686, "sha256": "597ef87ee25aee8492f980232f2bfa39e544d1f01b327589007c812c748f36b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/codex-inline-judge.json", "bytes": 3869, "sha256": "71551accd7b3bdc2019cfa8513d89a70e383685a04e343ff1dbe048eb4044131", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-tankcombat/tankcombat/tankcombat-qwen3-5-9b-q4-2026-10-21T22-02-21-878Z-rcrr/postmortem.md", "bytes": 10812, "sha256": "d648562175ae74a8d79d7e92256bd6d46ce3e9b55085483b7531cc89a77de687", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8.json", "bytes": 9604, "sha256": "8680de1a4b87d7d9130f1b1ed7de1dbab2760d023b80e97b7bdbb4bcaa4db6c5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8-desktop.png", "bytes": 5931, "sha256": "5e1c452f7ff64232d3a9a6e2f9274bc7a9696498404014ba1e1db5828191d302", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8-after-input.png", "bytes": 6190, "sha256": "3c4d2b2f99c4127f92d3af2bd520128cdf1a40122e804ec78f9a5d6c36f89b2a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8-natural-play.png", "bytes": 7458, "sha256": "647153e3603366abe7d75d9e85dfd670d6cad4c87a4aabf85d2b39831c577d05", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/ed363e56d8-mobile.png", "bytes": 4137, "sha256": "ebd4ab754889f883d6f50fc0a880d021b59be2fc7565e6b3d605d28d37a38f51", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/workspace/tank-combat-arcade/index.html", "bytes": 15943, "sha256": "01af0050f5a6a6b8585fe4e4d6c2d197c2c8a9edfae88aa09c7822eea4fe90ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/workspace/tank-combat-arcade/index.html", "bytes": 15943, "sha256": "01af0050f5a6a6b8585fe4e4d6c2d197c2c8a9edfae88aa09c7822eea4fe90ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/result.json", "bytes": 935, "sha256": "7c8ec5843337e7175cabdaa3e2ac4c7cc075d417e8f87505dbfe0d512f84f040", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/score.json", "bytes": 4448, "sha256": "12aba764fa61a8346f38fb19173f2cd2222667bbac1f7aead633364c0fc2e0f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/metrics.json", "bytes": 2387, "sha256": "1f5f4e43a7ddeaeb1e17330caa34d4400b50582ea670f37f17637652970e8109", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/facts.json", "bytes": 14953, "sha256": "085ec4bf9e384b43a86c74fbc9a1b9fb3c1a1ce08c30d286bc86560564359ca9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/codex-inline-judge.json", "bytes": 3410, "sha256": "8d01d427db73ac523539d07b2b42d2dc7c564407e41ba1bd25eec7e7603bc5be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-tankcombat/tankcombat/tankcombat-ornith1-5-9b-q4-2026-10-21T22-19-47-431Z-bk2x/postmortem.md", "bytes": 9545, "sha256": "bec7a08c91ee7537b0b00e6fb6a9e26fb588f11c9464b4689985e6d40f492bb3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/874af04527.json", "bytes": 5120, "sha256": "0d821acb354603f00c22dc198feb05ceebacefb8f7181d6fecfaa40019fe56c8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/874af04527-desktop.png", "bytes": 151505, "sha256": "30335a3d51225dc2589a008e89d612f4b6b83907e58dd992e38d967929ed8c30", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/874af04527-after-input.png", "bytes": 148452, "sha256": "daea75a16ea75244b68341f6fc3d463b0128b3b855749a5032b6adbb536473aa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/874af04527-natural-play.png", "bytes": 146654, "sha256": "a5d7c278fc006022489e09145a3f9834cfce8229d152aea3901756a7d819e2b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-browser-evidence/874af04527-mobile.png", "bytes": 27975, "sha256": "20edbb89ede8b3ec08b18ac19b0c3c4d34e6370ccfd55d0b85416f371c3191e1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/types.ts", "bytes": 362, "sha256": "ab0c7c1964e365a0228f598ae347d40e8a0880aeafa4435ae5477dda5a38e69e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/store.ts", "bytes": 783, "sha256": "9760695b5eb4ccd80b838c4b18b36999b073d9841b6cecb1523993efd4557824", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/handlers.ts", "bytes": 583, "sha256": "940f02f18a0bb6b8d2474c4a8e99244f58d6e36f9b17186a167c0fe41bf2404e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/migrate.ts", "bytes": 452, "sha256": "8463f4e06f92776f6912f17ff7b606ab79422ea83c91c7156ba212485f683c81", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/tests/migrate.test.ts", "bytes": 1297, "sha256": "8c33c5ddb272c7eaf7b88e31e8fd9c93f08beeb4acf0a0a1a4e8f9ee897090e5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/MIGRATION.md", "bytes": 3381, "sha256": "3e30b8efe8d85cf38d99ccdc86cc8a246e46e7c06298645375e11f8c652b90de", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/types.ts", "bytes": 362, "sha256": "ab0c7c1964e365a0228f598ae347d40e8a0880aeafa4435ae5477dda5a38e69e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/store.ts", "bytes": 783, "sha256": "9760695b5eb4ccd80b838c4b18b36999b073d9841b6cecb1523993efd4557824", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/handlers.ts", "bytes": 583, "sha256": "940f02f18a0bb6b8d2474c4a8e99244f58d6e36f9b17186a167c0fe41bf2404e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/src/migrate.ts", "bytes": 452, "sha256": "8463f4e06f92776f6912f17ff7b606ab79422ea83c91c7156ba212485f683c81", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/tests/migrate.test.ts", "bytes": 1297, "sha256": "8c33c5ddb272c7eaf7b88e31e8fd9c93f08beeb4acf0a0a1a4e8f9ee897090e5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/MIGRATION.md", "bytes": 3381, "sha256": "3e30b8efe8d85cf38d99ccdc86cc8a246e46e7c06298645375e11f8c652b90de", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/result.json", "bytes": 983, "sha256": "96dd4ee6a510c7f5d451d707a7b71dfba5797b8d549641a0a3b9cf0293a3ce46", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/score.json", "bytes": 2973, "sha256": "4b9d1ad39afb6f5784a8ec0f300adf4902527b1c5afb900460d9344a0ca737e5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/metrics.json", "bytes": 2406, "sha256": "a2ccb99ed0b77b5eefc6ba187f103257503ca13d67845051a5b00b5e89e851d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/facts.json", "bytes": 14905, "sha256": "f0eb5a4840a24ce5d3a1f5601f332c30edc59a62fd9c9bdee6dedc1a714f0db1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/codex-inline-judge.json", "bytes": 8843, "sha256": "22a3c715b5078d219994a824f4cb490b425c45f81461314062b783b65e6d77be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-schema-migration/schema-migration/schema-migration-underdog-saluki-27b-iq2-mix-2026-10-21T22-44-17-051Z-4t3x/postmortem.md", "bytes": 8993, "sha256": "2e5d96ca8b1c682589c6b3268e8f3f61aca06b92ede171b8a2f6ebad6d84939d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/1174d9dcd5.json", "bytes": 10844, "sha256": "c5a0e55dd8368d56b0091b0eb3e06f73f1c5dba19062499e0cf90ab054bf893d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/types.ts", "bytes": 433, "sha256": "427d85c960cd0388106b3439662c015d4e4dd775a44d4317f44d00b810e66afe", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/store.ts", "bytes": 848, "sha256": "fed0c675dd1566c88d6563bdd95fa8941c0f206d091bd8c309b7812bd391ef2f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/handlers.ts", "bytes": 650, "sha256": "5776c2ee2950a967ff79b2a726189db91c79db0f7846aed9f0f3bf12d28364b6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/migrate.ts", "bytes": 419, "sha256": "bfcd758aa870abaf4bab1c9d38de65e8c53eab661f638056db96c6c5a1cc9482", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/tests/migrate.test.ts", "bytes": 1193, "sha256": "f5b8839b16c583c28afd1d54244df5d8e55b41b289c3ab162e7ced81f1d63f3c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/MIGRATION.md", "bytes": 3205, "sha256": "96209ae65bafb3bbda86038c1627cf16a2b24ac020835d7182c9b57bb42cd3b3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/types.ts", "bytes": 433, "sha256": "427d85c960cd0388106b3439662c015d4e4dd775a44d4317f44d00b810e66afe", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/store.ts", "bytes": 848, "sha256": "fed0c675dd1566c88d6563bdd95fa8941c0f206d091bd8c309b7812bd391ef2f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/handlers.ts", "bytes": 650, "sha256": "5776c2ee2950a967ff79b2a726189db91c79db0f7846aed9f0f3bf12d28364b6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/src/migrate.ts", "bytes": 419, "sha256": "bfcd758aa870abaf4bab1c9d38de65e8c53eab661f638056db96c6c5a1cc9482", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/tests/migrate.test.ts", "bytes": 1193, "sha256": "f5b8839b16c583c28afd1d54244df5d8e55b41b289c3ab162e7ced81f1d63f3c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/MIGRATION.md", "bytes": 3205, "sha256": "96209ae65bafb3bbda86038c1627cf16a2b24ac020835d7182c9b57bb42cd3b3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/result.json", "bytes": 918, "sha256": "6b33ece190240227fdec35844f1fa2d577d8fc1b8623fb2348da85f43f4dbdb5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/score.json", "bytes": 2962, "sha256": "4c6ada182ceb691d77a2cf173269718d989a840b22e19b1fffa6dcdbb3ab97ab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/metrics.json", "bytes": 2379, "sha256": "795581cc2a963a7267cbb9dbecef4791a253515e56fabc7d1f91b33d353842bc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/facts.json", "bytes": 14763, "sha256": "c02d7b6cba6c51d267eb6c752eb269407986b42a2c19fda700bfc4a50d8db97e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/codex-inline-judge.json", "bytes": 8325, "sha256": "9de4df52bb8e23e3c05fe70e002458271db64065363a509a501c60306626a9ca", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-schema-migration/schema-migration/schema-migration-qwen3-8-27b-q2-2026-10-21T22-51-36-578Z-1asz/postmortem.md", "bytes": 9049, "sha256": "fe6e8c1f4319e01965420bc80006d0e139f2d78ae00928fe0e48a48d4921162a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/e232744039.json", "bytes": 10977, "sha256": "dcd485ee0064139f3f1e5c9822226bd46338dda5eb9e66c85ac11dec4fbe0393", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/migrate.ts", "bytes": 274, "sha256": "4476a13c1d644a3a2566fbe1cba960bbdef2394b313830daa970472eb9ed7830", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/tests/migrate.test.ts", "bytes": 1068, "sha256": "682d425e99ba7e55499ebd12db15369a1fdac995f12e0f952168ee64bea02491", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/MIGRATION.md", "bytes": 2888, "sha256": "3ee34593f793f91247f908d25b672d6d95602d92e13a05375035bc26b8baae26", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/src/migrate.ts", "bytes": 274, "sha256": "4476a13c1d644a3a2566fbe1cba960bbdef2394b313830daa970472eb9ed7830", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/tests/migrate.test.ts", "bytes": 1068, "sha256": "682d425e99ba7e55499ebd12db15369a1fdac995f12e0f952168ee64bea02491", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/MIGRATION.md", "bytes": 2888, "sha256": "3ee34593f793f91247f908d25b672d6d95602d92e13a05375035bc26b8baae26", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/result.json", "bytes": 913, "sha256": "5026d4dcfe927b550783c17003da69a763b1da7a734af86655a94432123c044e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/score.json", "bytes": 2945, "sha256": "68f150771b8be014b6331b13dd6c82f13e1c37c1f5cf89454138ffb71e0f3a2f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/metrics.json", "bytes": 2385, "sha256": "bed0110c714fb5f753a1d85f47ca5a06fd665fd219ad243643e9abcb5b328ed8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/facts.json", "bytes": 15679, "sha256": "a970fb405c36130c434237452f697736c6565432589acb1fe05ce6d2cb9c4b6c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/codex-inline-judge.json", "bytes": 8279, "sha256": "67f6863b0886ddeefa769ec17a3e0424a30120564e082b615bbd965f05f6615a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-12b-q4-2026-10-21T23-03-19-913Z-ejny/postmortem.md", "bytes": 8944, "sha256": "df8637fa1597c956de21c82959040d1b5e1e28239056c3c639da78e2c0858ff3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/0732c7ad7e.json", "bytes": 10270, "sha256": "cc05ffed157241873d5d1c453f03afe798f3691c060ffd960823449440dce20a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/types.ts", "bytes": 193, "sha256": "0c9a266194352ea16117ff3b7dc5471f7548d9eca2b1f6dbe5c9d7564ba136ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/store.ts", "bytes": 1094, "sha256": "e9246d7faa1ef4f4be8a34fb3aa13921911b04115b09ff4f2a4393b4c75270e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/migrate.ts", "bytes": 974, "sha256": "bbb22d8469de56b3c2af40914d3ee86fd39686cd2db53201d50a76030432dc4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/tests/migrate.test.ts", "bytes": 2188, "sha256": "9c2db135b4ee6334cbab2acbc8401d5ee9ac61889266c95e4f68d0eb6e107799", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/MIGRATION.md", "bytes": 4290, "sha256": "ab6b73e166bec0ab68a4a0a45ca5ed1cc38c2eaa680a38606103174167482fcc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/types.ts", "bytes": 193, "sha256": "0c9a266194352ea16117ff3b7dc5471f7548d9eca2b1f6dbe5c9d7564ba136ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/store.ts", "bytes": 1094, "sha256": "e9246d7faa1ef4f4be8a34fb3aa13921911b04115b09ff4f2a4393b4c75270e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/src/migrate.ts", "bytes": 974, "sha256": "bbb22d8469de56b3c2af40914d3ee86fd39686cd2db53201d50a76030432dc4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/tests/migrate.test.ts", "bytes": 2188, "sha256": "9c2db135b4ee6334cbab2acbc8401d5ee9ac61889266c95e4f68d0eb6e107799", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/MIGRATION.md", "bytes": 4290, "sha256": "ab6b73e166bec0ab68a4a0a45ca5ed1cc38c2eaa680a38606103174167482fcc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/result.json", "bytes": 912, "sha256": "7e7ac4b1fa842cbd8427daeeb78042a819d3e83f88be6bc5eea185b513bf7a1e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/score.json", "bytes": 2961, "sha256": "bd93315f31e3e7bbdde06cf3de7b0c32b97f533f1d33f8723ab5a02e4a77d7d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/metrics.json", "bytes": 2382, "sha256": "08adea7a597db7f64d179ea2d2d20d5007c6556010c7182e7d86c24584c4d22f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/facts.json", "bytes": 17466, "sha256": "9f93489251f807423325650a386caa5ab0c7f4284914fb09d85c99fb6dc9e5ac", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/codex-inline-judge.json", "bytes": 8290, "sha256": "3a5bb158c9f660751ab062026b3fb4314a88d12ca9be406cee59b4164d5d78a5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-schema-migration/schema-migration/schema-migration-gemma4-e4b-q4-2026-10-21T23-06-42-982Z-bz8m/postmortem.md", "bytes": 9035, "sha256": "ac8d1f9309010656504c995a9ef96de9a86d780973878c2a73d8679d63c1e388", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9773a858d9.json", "bytes": 10272, "sha256": "578c8f4004754698e8c18f7e9476c925084f31534cf2d2f06ad1982061fbda0b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/types.ts", "bytes": 368, "sha256": "d0d6204b766174b04979dd599ec01f834718b3c4ad592ef7ac252bbcc6723dc0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/store.ts", "bytes": 1088, "sha256": "c8703cee25f73689571e8e104dff4de99539c1a1b2a7bccaf1fd94e485edc9fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/migrate.ts", "bytes": 442, "sha256": "3220093fd8d2dd8556b122fac59bbb2c2d2842c83a5a1eea1fff080c7141d65f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/tests/migrate.test.ts", "bytes": 1285, "sha256": "df825da0d8ef8ddc5bff606051e206866a0d8f1390d036ae049eda1843cebd94", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/MIGRATION.md", "bytes": 3193, "sha256": "a6e5d26e06af208d2c99107643287bf859385fd789bcba978811cd8632140543", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/types.ts", "bytes": 368, "sha256": "d0d6204b766174b04979dd599ec01f834718b3c4ad592ef7ac252bbcc6723dc0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/store.ts", "bytes": 1088, "sha256": "c8703cee25f73689571e8e104dff4de99539c1a1b2a7bccaf1fd94e485edc9fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/src/migrate.ts", "bytes": 442, "sha256": "3220093fd8d2dd8556b122fac59bbb2c2d2842c83a5a1eea1fff080c7141d65f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/tests/migrate.test.ts", "bytes": 1285, "sha256": "df825da0d8ef8ddc5bff606051e206866a0d8f1390d036ae049eda1843cebd94", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/MIGRATION.md", "bytes": 3193, "sha256": "a6e5d26e06af208d2c99107643287bf859385fd789bcba978811cd8632140543", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/result.json", "bytes": 912, "sha256": "f377cea9aa891183f7627bc34857c1edf10af1f817635ee424ba34bef794a4a5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/score.json", "bytes": 2961, "sha256": "3d46518841ad4b614463179087b1bb19c5452da24735747bbbf2fe6e2d3d4b00", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/metrics.json", "bytes": 2337, "sha256": "268e5a26f4b91754ae7256affbe43b84a72e93100198a069b68c9d5bf5e84055", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/facts.json", "bytes": 16371, "sha256": "a34250709480ca560cfbee98f06da5e488e0ec4396ac2f909cf0457ec3f93aa4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/codex-inline-judge.json", "bytes": 8313, "sha256": "8929b493f7c657e9c55d2a56f8cc5e3a1d1c6145d85b86fa2205de0b38735748", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-9b-q4-2026-10-21T23-10-03-131Z-nm8n/postmortem.md", "bytes": 9069, "sha256": "a2ee482f0b3d8df69a51434c1d06e0a17ed8aa6235d7a680cc115deaf0219f13", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/128ff317ff.json", "bytes": 10271, "sha256": "06fb3fdd854c9160040b71e2e4cce6484760557fa5b117d1fadb3a9dd333ee6c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/types.ts", "bytes": 433, "sha256": "e2a15704790e0ec60ceebf3ab874b5824678328fdceef738e682416dce131629", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/store.ts", "bytes": 704, "sha256": "4a7cfda41381920fe06feeb7d5ad9d91d907abc60b0a478dc6d9a870f8bf5378", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/migrate.ts", "bytes": 910, "sha256": "c1d8f0781dc80f521b230fbd26d7822722b047e517730046c49daa5301ad1725", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/tests/migrate.test.ts", "bytes": 1144, "sha256": "a8e2a80c277df5a77014b5e0257c032616193e5006fe0ebac0ec687e4b2eab4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/MIGRATION.md", "bytes": 3032, "sha256": "6ece9d15501ef9eb07830ecfa055a1649541c6ed34c54262a53bbb940736ca94", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/types.ts", "bytes": 433, "sha256": "e2a15704790e0ec60ceebf3ab874b5824678328fdceef738e682416dce131629", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/store.ts", "bytes": 704, "sha256": "4a7cfda41381920fe06feeb7d5ad9d91d907abc60b0a478dc6d9a870f8bf5378", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/src/migrate.ts", "bytes": 910, "sha256": "c1d8f0781dc80f521b230fbd26d7822722b047e517730046c49daa5301ad1725", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/tests/migrate.test.ts", "bytes": 1144, "sha256": "a8e2a80c277df5a77014b5e0257c032616193e5006fe0ebac0ec687e4b2eab4b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/MIGRATION.md", "bytes": 3032, "sha256": "6ece9d15501ef9eb07830ecfa055a1649541c6ed34c54262a53bbb940736ca94", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/result.json", "bytes": 922, "sha256": "17dd669844726223a3a29b9516c3a02ae89ed339663876f10af1aaa2f53b0174", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/score.json", "bytes": 2965, "sha256": "6d6ed51ba7c59f96cf63fae230a5905635d47f4e8674da34c7c7de21d4e27ea8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/metrics.json", "bytes": 2344, "sha256": "7be07f7e1720c233c19bf37d136075bbf2ba79ebc268744822535f75eb1fc1fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/facts.json", "bytes": 16706, "sha256": "6c509b238360a33b505fa0f7aa86ed1f84c34a3d1ef92a1bdad9fb3262ab8fd1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/codex-inline-judge.json", "bytes": 8372, "sha256": "0e6f2f9c9448b739a3971be8d765377952204058c363b19b76938d15835f9427", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-schema-migration/schema-migration/schema-migration-ornith1-5-9b-q4-2026-10-21T23-13-14-190Z-ieyb/postmortem.md", "bytes": 9044, "sha256": "dbfabdcc227a0c1e10310463c32a45bb5ef69eae2f0bb7bc864e85957ddd7e41", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/dc4a2129a8.json", "bytes": 10352, "sha256": "a51a04d9efd715f268912f30277402973eca1331e7730510d4819d1519888f5e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "32ed67b8728ed6a20fd6434d95ee93e16921b68685d98cca60465f3b43414bd6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/handlers.ts", "bytes": 624, "sha256": "ea886137ca6d3028892fa1f67270b25f04fd425710cdd6ccdbda6249256d3e75", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/migrate.ts", "bytes": 335, "sha256": "dfebad0bf30fc48b4eb09f8fb95ba64fca5afcb64e182cc508b5443631b7c6da", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/types.ts", "bytes": 194, "sha256": "1227b8e1014c44b86ea048962fb4659aebd34e941e8d08a8087f9150a599eec3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "32ed67b8728ed6a20fd6434d95ee93e16921b68685d98cca60465f3b43414bd6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/handlers.ts", "bytes": 624, "sha256": "ea886137ca6d3028892fa1f67270b25f04fd425710cdd6ccdbda6249256d3e75", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/src/migrate.ts", "bytes": 335, "sha256": "dfebad0bf30fc48b4eb09f8fb95ba64fca5afcb64e182cc508b5443631b7c6da", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/result.json", "bytes": 1635, "sha256": "7eed659d1e2b8d11669b0ff79a77cd6a6676100abb1e9d5aa4f6fa74aa2bf59d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/score.json", "bytes": 4545, "sha256": "7b6cd4129208c056a9098454a1eb55d911c7858b7da2fe2836c0cd9d60242724", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/metrics.json", "bytes": 2368, "sha256": "39bbe31d9b2becb4f2644bc5b307b6c382dc21cd162d6db5f0198d2e47270843", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/facts.json", "bytes": 16899, "sha256": "1e37c1d3c1e250a6c511fe9ef8cfda39d2cfe926c61e6f99da7c725015fdf992", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/codex-inline-judge.json", "bytes": 7077, "sha256": "d896d5977390e95aadbca41ce8518625253a3941c3ebdaf3da69711eb1c190f0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-schema-migration/schema-migration/schema-migration-qwen3-5-4b-q4-2026-10-21T23-16-51-850Z-g5i4/postmortem.md", "bytes": 10660, "sha256": "3ce6a99368221d2bbaf3121c37446d7096ac6c2365b7148296cd333b038a7bc6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/be5486180b.json", "bytes": 9528, "sha256": "a461a3e13339d88dc70cdc63d4ef9a0786683c537941745146fbf09871dc9ea6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/src/machine.ts", "bytes": 1657, "sha256": "84308c5f11fb0f2725452f22ec86602b9fdd2d71735a9b2953612ae90f1bd5ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/src/machine.ts", "bytes": 1657, "sha256": "84308c5f11fb0f2725452f22ec86602b9fdd2d71735a9b2953612ae90f1bd5ac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/result.json", "bytes": 908, "sha256": "71e06e5783d92669925250a5fde59fe2b43d25d1469a4f90ddc634b6df3e09ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/score.json", "bytes": 2855, "sha256": "bae10594e54526b9c4668da08fd76c42526fef76f21dd60611f488bb5a352f12", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/metrics.json", "bytes": 2364, "sha256": "d08e6ebc14d254660918b8ae9451dcbdda084adceb0679f17680bf8ee39cc9ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/facts.json", "bytes": 13661, "sha256": "f57db51ad49f9e6d14a0cd5dfd746269ee9802b5dd411b3cdd76b81dd2cdac59", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/codex-inline-judge.json", "bytes": 4906, "sha256": "fd32f2d464eb3c0d25c0a226a363d4310cdc63932300a495b8fca30037530015", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-8-27b-q2-2026-10-21T23-18-58-376Z-wbi1/postmortem.md", "bytes": 8802, "sha256": "2e83f8465d905a52a91c67858afae9428c7ebfda3c031c04d96c605c6732fb1a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/e8824696b4.json", "bytes": 45527, "sha256": "7760d42c2f6a8618a4c9b77bf5f5d234e13a0e4807c4857a4968039fc5108405", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/src/machine.ts", "bytes": 2214, "sha256": "ecac25d527e759ae64664dfbf526bf33a2ea1db3675359d0f57138171ebc6df8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/src/machine.ts", "bytes": 2214, "sha256": "ecac25d527e759ae64664dfbf526bf33a2ea1db3675359d0f57138171ebc6df8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/result.json", "bytes": 902, "sha256": "48fdce4da6796661d8f7663f90d6f59282160bfa4639f2b1a4c56b400cedf20e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/score.json", "bytes": 2838, "sha256": "9c83dcfed6d9defd312b5ff4a96993b774920ed0793c7ccfab3ffe230719aa54", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/metrics.json", "bytes": 2316, "sha256": "4f693f7553fef1f99f99a3dd0c1d7469bcfc581690ad139b62eac3d6b4420233", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/facts.json", "bytes": 11986, "sha256": "efaec0661d775ec5440afa4f4d05b2a3fdef2af8c9672140cf3fd20317ce9916", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/codex-inline-judge.json", "bytes": 4908, "sha256": "26fea6613633698edc62d45f00c03ffc19b354c88ec44c9e4d334607a12c6122", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-12b-q4-2026-10-21T23-24-43-984Z-s7gm/postmortem.md", "bytes": 8790, "sha256": "7e0e1a75f771abad169deac15f08a6f32d2f36673b4184921351967b2eedefae", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/e95d204c40.json", "bytes": 44584, "sha256": "c97264213df4780bdb0b0169519a01cdbec2ee973d52e1b118a3f7f80576438c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/src/machine.ts", "bytes": 3227, "sha256": "69c406c5926c94010569999bee01b5a3855636617bdb7d0dbb0a4e68a8573327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/src/machine.ts", "bytes": 3227, "sha256": "69c406c5926c94010569999bee01b5a3855636617bdb7d0dbb0a4e68a8573327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/result.json", "bytes": 902, "sha256": "dc5f70545e56309e587905c47c0ccbef7a7a4915805fd0016adf40adcdb450d8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/score.json", "bytes": 2834, "sha256": "de7203772d839ba5a5534b7f815898a1eafe1268ffecbf36b1dcc13333b90d1b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/metrics.json", "bytes": 2391, "sha256": "4fd6f447b1cb6f6f5a0ca17099ca4ef8bfa96481e4be61423210482a7079cfcb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/facts.json", "bytes": 16307, "sha256": "e1c534f3815d397c8f2ae1acca960a5acaac4e29af85e6df34ed2812db019c5c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/codex-inline-judge.json", "bytes": 4939, "sha256": "1b5c02adeebb7bd9edc32a398b712b07dbe1c4cfee01cc0d01f2bb97c57ebd1c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-gemma4-e4b-q4-2026-10-21T23-26-12-649Z-zkhh/postmortem.md", "bytes": 8854, "sha256": "e46d04b92719b45a2b0a865d11db30c32d1608a0474f05eb4b03cf8d60b632fb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/adc1f30dc0.json", "bytes": 47861, "sha256": "43a2c0b4f1d73a239382431c721827b276d5c3798b29ed3fe4582d14a399655f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/src/machine.ts", "bytes": 1890, "sha256": "ab75aa519543f224515f6b4744ec1efd2789495cdaf84613c7de76c9fd5b5387", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/src/machine.ts", "bytes": 1890, "sha256": "ab75aa519543f224515f6b4744ec1efd2789495cdaf84613c7de76c9fd5b5387", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/result.json", "bytes": 902, "sha256": "9b592e1a9f54c6acbb302280a0d160523545661d04afdf9cf3f39800dced565c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/score.json", "bytes": 2847, "sha256": "aa66dd0c94edfa11ad3aac75bb0477ad9da9d6695ba8c05ca317b9527c375219", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/metrics.json", "bytes": 2334, "sha256": "ebc8bcf1618d06fd52b86655edbd168d63dea9049c66ab74dfec38fc9355ef2c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/facts.json", "bytes": 16908, "sha256": "ca2ed429779658e48aaf2d135848c3eed0ea96eae9e5d3f8c7967bf70393dd48", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/codex-inline-judge.json", "bytes": 4897, "sha256": "372a68feff44c4452004875b3620961b6b4f29a6c87a93beea59e6b58a0a1cd9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-9b-q4-2026-10-21T23-36-01-248Z-dyer/postmortem.md", "bytes": 8790, "sha256": "d2d1b302d19b2deacc3b98b6a1d06c31878c1bad303688ffae32a33babfab27f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/3e41108a46.json", "bytes": 45508, "sha256": "31aa3624f4fd0220306075c533ed7f68b475fbdf38fc39f8dccb97ecf88c79eb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/src/machine.ts", "bytes": 1358, "sha256": "4010eb83cadc0638b4dea93f0d1b2d5640bdde61fa33bac06e5bcf6e6188d632", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/d82ea8785b-tools.jsonl", "bytes": 3473, "sha256": "632292d9876d8cb11ce78bdea35c982f493d74ad2a2b73f193cd538b74147a81", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/d82ea8785b-runtime-log.txt", "bytes": 2434, "sha256": "e659ee7e1b151523a103112428f9e73a5327372294b59fa89e8760393cfb7eb3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/src/machine.ts", "bytes": 1358, "sha256": "4010eb83cadc0638b4dea93f0d1b2d5640bdde61fa33bac06e5bcf6e6188d632", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/result.json", "bytes": 1805, "sha256": "f2475e7e37dff9a18414fc1dcac7edaf1a012d5125a648bfe7a3270a13a76e80", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/score.json", "bytes": 5156, "sha256": "a363d7c6ae5bc0da4cea85ddddc41ba387cfbaa1349401f45d920aa817e3c739", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/metrics.json", "bytes": 2370, "sha256": "bdc5bf4ef0a3c9d11d8f250c897442c06e44ba70a8721602c2aa809168c08a6b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/facts.json", "bytes": 16829, "sha256": "8df73982e26052d9f4744a3d8e745edddc2225f42e492a5c0c05e2298585d1c3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/codex-inline-judge.json", "bytes": 6051, "sha256": "e86189625acb8a5f66e2e6e033229fd2421765335523cbd35c765cf843009c65", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-ornith1-5-9b-q4-2026-10-21T23-38-53-723Z-b3l7/postmortem.md", "bytes": 10806, "sha256": "f346ed432ab495fd46dd748b10b7b1e43294fb85aaca4967aae4a536e6922a65", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/d82ea8785b.json", "bytes": 27253, "sha256": "e391983f85e4866ab805da9f24ea9b25f10083dd72318eef431a38c0913e4a89", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/src/machine.ts", "bytes": 1981, "sha256": "f8de8d3c3cd4a10fd9993ba8098ce7f5e0334a7b2b258e466c7dc8026c161778", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/src/machine.ts", "bytes": 1981, "sha256": "f8de8d3c3cd4a10fd9993ba8098ce7f5e0334a7b2b258e466c7dc8026c161778", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/result.json", "bytes": 900, "sha256": "206420d77e2c4e69f57931fc14b4e357fb3496c8933433a272921314b871d46a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/score.json", "bytes": 2839, "sha256": "c9e82c31ed422209960e28321683754ba28b92fe7ef46d0dd0bf4228c37231a7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/metrics.json", "bytes": 2315, "sha256": "cdd55ad947960688c9a0b04f0a30a4909ddf0d46d2e6582e0c824f04af950fa5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/facts.json", "bytes": 12135, "sha256": "8d68af55bd4ff79aeb162c4060d6d9959bd716caa9564d7a2d5a965b5ec57ec8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/codex-inline-judge.json", "bytes": 4935, "sha256": "9c7dedccd6257f713d687d32fd4553c289d2819327e86ab978a0bf1c5dd00fc7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-qwen3-5-4b-q4-2026-10-21T23-42-35-115Z-tily/postmortem.md", "bytes": 8851, "sha256": "d1d274c329771254e8b09bf343a868a7e57366e4103a16dcfc2e522395f8cf25", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/babbbf2f7c.json", "bytes": 44551, "sha256": "f865086fc58638eeb0821d360580d252f3840da015f3539bd4ee0e2c5986d3d7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/src/machine.ts", "bytes": 1598, "sha256": "2cd158b31bfd54e33daad2a05c8055024f13b977ae24cd797ca444b7574b1acf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/src/machine.ts", "bytes": 1598, "sha256": "2cd158b31bfd54e33daad2a05c8055024f13b977ae24cd797ca444b7574b1acf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/result.json", "bytes": 973, "sha256": "47d822159c5a2d92864384ba54dc2d88fd629c784a5cafc27d2cf3dd235e82ab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/score.json", "bytes": 2880, "sha256": "49a09590c11bb4749f85f1d08d4fd472a54c1282f0536c225ed8f10d8fc7d426", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/metrics.json", "bytes": 2345, "sha256": "42174f13466a417f8749cda1c3284a4a5404ef13f70ebebbf2c5831f96b3661a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/facts.json", "bytes": 12158, "sha256": "2b95cb4134480b26a00ebde91204a5e21bbf820081d058b33260797c18a49dbd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/codex-inline-judge.json", "bytes": 5028, "sha256": "ea8ddacf7145c6aabaafa8de373ac0db03bd67c7c1e417d495d227eeb852115c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-underdog-saluki-27b-iq2-mix-2026-10-21T23-44-14-996Z-y2pk/postmortem.md", "bytes": 9480, "sha256": "b6bc2d522db96f7c9170a34bd58d7444d9a537deb4c6661dbad56a7b3879e20f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/7884b5e425.json", "bytes": 45774, "sha256": "6c8b2052038e9c02bf4295710b498acb6e0fdffb4da8fc339c0331d49e92a38f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/workspace/pagination-service/lib/paginate.mjs", "bytes": 1185, "sha256": "50bcb25929f0ed884105583949391add054e15d0a3e7f1c8713a3a1e2131883c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/workspace/pagination-service/lib/paginate.mjs", "bytes": 1185, "sha256": "50bcb25929f0ed884105583949391add054e15d0a3e7f1c8713a3a1e2131883c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/result.json", "bytes": 864, "sha256": "ac7a14e5108b1c6324dc02af7de9f30c95534998e0069c51039e74f26410e402", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/score.json", "bytes": 2807, "sha256": "3f50f65b1529fb35af107b6f6b75db16a76254c141004d5c69d8751214e35820", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/metrics.json", "bytes": 1305, "sha256": "7dbd9c22b3c3e1dc1b44358166d54b6b89476db3cf02180044ea57180a6b67c7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/facts.json", "bytes": 10617, "sha256": "53760f9350edd9a832b0a2ff478eb62abc3f6ad114d90ddcde4b47defed50a4d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/codex-inline-judge.json", "bytes": 3854, "sha256": "97969cc1f31bd2e53beb73855a769e429adfb39228f0214803f1b5a029fed8cb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-12b-q4-2026-10-21T23-46-53-261Z-owhv/postmortem.md", "bytes": 8986, "sha256": "8cb2b30ad55654b2e5cc0e535048a3b778441c00694a3871adaff92eb5b048d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/2d56e46bf3.json", "bytes": 9705, "sha256": "268cadcf0898f43c4802e205a6d5a6663f5fdca86db42b1238ecb6788576c754", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/workspace/pagination-service/lib/paginate.mjs", "bytes": 1187, "sha256": "648c72b08714655c84925eedc78a430254ad3b2d64bde9cda8a2a35df300e43f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/workspace/pagination-service/lib/paginate.mjs", "bytes": 1187, "sha256": "648c72b08714655c84925eedc78a430254ad3b2d64bde9cda8a2a35df300e43f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/result.json", "bytes": 863, "sha256": "f7782108a38dbc83fc7f1115362fe4505a35b593fc3ebd329fa26e39a60ee2b4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/score.json", "bytes": 2808, "sha256": "1a37cb0d32cd8f6b4161c27f44221cbc3201c43e96adc24681887c059c5d4dfc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/metrics.json", "bytes": 1304, "sha256": "29b0e01d04a00aca438d0b6e9ec2591e628ce63ed565556485cd9f99f6459fe1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/facts.json", "bytes": 10449, "sha256": "0a3cf4509e8591a633cd75d734d04f85e7c2251ff13dd00234016d308702029d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/codex-inline-judge.json", "bytes": 3835, "sha256": "2d34dee053c83cd73e8f9ebf6ab5cf5f6c725ff014cb8321a29136cda82e5a53", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-symptom-debug/symptom-debug/symptom-debug-gemma4-e4b-q4-2026-10-21T23-48-57-030Z-lxx7/postmortem.md", "bytes": 8933, "sha256": "90975da09f92b5487bd0e321b9cf326254031c6c0deaa0f93b4e12d873388415", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/28991eb661.json", "bytes": 9649, "sha256": "388bdbf33df409162c0636425231418b39ffea8bcc76c925f0435fedba3f0531", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/result.json", "bytes": 863, "sha256": "5aeee85bfecb241f26b1192fa0aeaffb8a60e7803bf21150332620bbeac01852", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/score.json", "bytes": 2808, "sha256": "aa10b162dcbecbd17e245d407d1a32a0c967354efec27b2a0bed514fd36a8fd3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/metrics.json", "bytes": 1303, "sha256": "670fed33f556236339588bb54f4ed433d31d34c09bf1ecd6a2bee49084df8072", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/facts.json", "bytes": 10449, "sha256": "83844ec8690928267ba56a2cafeef10b904342e74ad33ecf396ac8f41393a33a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/codex-inline-judge.json", "bytes": 3851, "sha256": "5dc85844524e970ba23eea0aaa5fbf47892fe4e566a573dd6fed0ad2b46daca1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-9b-q4-2026-10-21T23-50-28-314Z-zay9/postmortem.md", "bytes": 8962, "sha256": "162591d323c6e9d110e3d04e55ffafc37cc5f666bc6b96e8e05dea3e95b05d89", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/35723d5cb4.json", "bytes": 9649, "sha256": "01a269baff8c0ec0e98f549a840149f7b4104aa15e4e64d81e6855d63e82e846", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/result.json", "bytes": 873, "sha256": "d6e450e4789c4f8b13f4b27e8ceab801c2c7a8f4f603564bfddaaf07eaeeaeba", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/score.json", "bytes": 2811, "sha256": "726722315c59de0b88950600a52e30f565a8a4f456692ae5f386801735226acd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/metrics.json", "bytes": 1306, "sha256": "76fcdb00fc9418c1d843be509ad83bff53841f18f30ff27888a7c134247d8126", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/facts.json", "bytes": 10508, "sha256": "0c0954144aa601433422f7769ab18bb939b10dfafcf17e8e15f377eeab684512", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/codex-inline-judge.json", "bytes": 3948, "sha256": "9a91de119001274bd1a61fb4a5beee15ccfce3909faba93ca439ebf5a7e05beb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-symptom-debug/symptom-debug/symptom-debug-ornith1-5-9b-q4-2026-10-21T23-52-12-932Z-tu9u/postmortem.md", "bytes": 8733, "sha256": "2b1e2379976178ac03147150e546fa5d70dbe164c5fefc4c99ae98babbb1b981", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/9843bc717e.json", "bytes": 9710, "sha256": "75ed3d13befe7bb9cdda703fbe7356a2c7fb19a80ed5bf91f3a877b855ef8426", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/result.json", "bytes": 862, "sha256": "110ffcd55e2dffe67ce77031bc6fe1060c218b6b3180de2ead190cf68c847bde", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/score.json", "bytes": 2804, "sha256": "9945baa4f77f4f07c4b9b5012cc9d0fa37c566f9b7fd07bc0cdbedfe5c270e66", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/metrics.json", "bytes": 2368, "sha256": "ab799199bee21959337264197f99d887f55d3b011de9e32e0dd293b4b817a934", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/facts.json", "bytes": 14047, "sha256": "f5826ed0beef1d7ba11411b16e7ff3bd89155690d80217f0a42ca17a725533cf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/codex-inline-judge.json", "bytes": 3924, "sha256": "48b4186c3e6850e648bdadfa25799b24b4bd0ef58f51a3a8c08314c4faab1820", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-symptom-debug/symptom-debug/symptom-debug-qwen3-5-4b-q4-2026-10-21T23-54-20-831Z-infk/postmortem.md", "bytes": 8712, "sha256": "b0baaac98339e3efa41c667792cf6a9f570dd611d14bf801145561ce9c594244", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/beeea1ff13.json", "bytes": 9734, "sha256": "e7d0f81ee86854c3546fe9567e3a4f8918c7f2d5ca2b09bed2817dde0f2615a9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/workspace/pagination-service/lib/paginate.mjs", "bytes": 1207, "sha256": "d266df72c2209ceb388c9698ee651e2c8c6ac56d7a6fe9dd1b2a3e4577c51bce", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/workspace/pagination-service/lib/paginate.mjs", "bytes": 1207, "sha256": "d266df72c2209ceb388c9698ee651e2c8c6ac56d7a6fe9dd1b2a3e4577c51bce", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/result.json", "bytes": 935, "sha256": "d8f5f9ac6aae08dee23555b59a6d23fe071c7608dab410c9b9f9e828dd810b19", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/score.json", "bytes": 2838, "sha256": "ce0337d472660d51396a88d13fbdbfe1e0fd400fc352fe0af997ad406cc05aae", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/metrics.json", "bytes": 1310, "sha256": "74bf862e5f7d7a461063dadf4a4f290636f04aed9f107ec91320252517425f07", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/facts.json", "bytes": 10717, "sha256": "e63cc72548bbb90c9e62d1184dd8481e95e8234019f61fd05af0a86b2af89339", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/codex-inline-judge.json", "bytes": 4123, "sha256": "222ba6c5ee8d5e6d535ed97ff2576c29848e270b38c03fafd5cef050b4930fca", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-symptom-debug/symptom-debug/symptom-debug-underdog-saluki-27b-iq2-mix-2026-10-21T23-56-33-611Z-ikm1/postmortem.md", "bytes": 8869, "sha256": "069afea3bcb5c46878d63f9a40a5ebfd1d083ba026438fe30b66939705bc3de4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/5e1899dfc5.json", "bytes": 9833, "sha256": "202ffd80ff08ab73d96a18d91da827f48db8d70db8386be8375e06a0f7da91ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/workspace/pagination-service/lib/paginate.mjs", "bytes": 1211, "sha256": "5860914bdd874591b431f3ebd3e80b66d82f0837aa48be81aa928a313f34e166", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/workspace/pagination-service/lib/paginate.mjs", "bytes": 1211, "sha256": "5860914bdd874591b431f3ebd3e80b66d82f0837aa48be81aa928a313f34e166", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/result.json", "bytes": 870, "sha256": "a7e35f22d7e27feba039bfbbf2fd1ce6a73f3b4899f744ce8c18492055a34985", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/score.json", "bytes": 2812, "sha256": "98a2e4914bbcef9d61b718a7e42cbf7900c74cc9e634c63621f053bc06188d2e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/metrics.json", "bytes": 1307, "sha256": "47900b3659c327f56f62e5bf2a684978ad42a0218a128e58291ea2eb8595b3b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/facts.json", "bytes": 10582, "sha256": "6cef29e49525419b94ea9dec17f5d0c3c3e0d3bb215ee4c2382b7faf7324eeb6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/codex-inline-judge.json", "bytes": 3967, "sha256": "2f16d71f4448c195c2240f1a1bc6c9df9cba0faaba96f8531511233e6f63723a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-symptom-debug/symptom-debug/symptom-debug-qwen3-8-27b-q2-2026-10-22T00-01-51-457Z-kmkw/postmortem.md", "bytes": 8812, "sha256": "a53d43b9861a15ed1e7999696d49612a81b33e4af92b96c7fc4071b3e4de3626", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-code-evidence/72de12e1d3.json", "bytes": 9664, "sha256": "ed798cd2166b32f5cc5f691b47482d94b6cb559a1e490b78f76ed1ac5d94f96f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/out/customers.json", "bytes": 5448, "sha256": "24bfe2110873ab95aa42398d6205a7a18af80714997d81554567fa415f29569f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 5863, "sha256": "616dfb29843f4c49420322267016c2a96c49703b3f9c9a4d5e0c4d9c87624a4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8de5cf5028-tools.jsonl", "bytes": 1159, "sha256": "30d346e2b9d70dba4e8ee650a3d17ed5d35aefee7eac4c0d164eda94fae399ad", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8de5cf5028-runtime-log.txt", "bytes": 325, "sha256": "8470ec93c32e380786797aa3abcc3f0f2c5ff33dd69822e1831d1cebd6f46ca0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/out/customers.json", "bytes": 5448, "sha256": "24bfe2110873ab95aa42398d6205a7a18af80714997d81554567fa415f29569f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 5863, "sha256": "616dfb29843f4c49420322267016c2a96c49703b3f9c9a4d5e0c4d9c87624a4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/result.json", "bytes": 1278, "sha256": "06e0bcebe2d240b252f8bd3ca8fad3c4ffdfc08a8a2524eeac8924ae8ac14e22", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/score.json", "bytes": 4154, "sha256": "df0d1fedf8e9125c52fa56e50bc3482ad26959a29b6c8e335a02a99089edfde2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/metrics.json", "bytes": 2382, "sha256": "da4104fe13042ef7049683c109078b138988884a0ae3479af708d0bfa6348997", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/facts.json", "bytes": 13571, "sha256": "e4a788b702ca95e4632b915a7e7c682421f8a5fca7a2177c62f91aaa40b0b301", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/codex-inline-judge.json", "bytes": 7135, "sha256": "bdda4ec3a3e5c989e26f1a2b837a99d515ed6c429fbad6398fe93fcd17b9533c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-e4b-q4-2026-10-22T00-06-24-259Z-cipq/postmortem.md", "bytes": 11161, "sha256": "4b2ae3a986f9044e9c1864952fea9c7d61bc8a80236a1a5a5fac457df8f1d3ad", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8de5cf5028.json", "bytes": 10256, "sha256": "ad97a693ef18c5b52d39593b9d2dbde8b85bec413af215b3d4c7310b2a657213", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/out/customers.json", "bytes": 6738, "sha256": "55741ad10f83e45d112bc72a61142baafe03cf2c5f0ac3603ae66ceb4a073d6e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/scripts/clean_customers.mjs", "bytes": 3133, "sha256": "6bd0bef50eddeb9bd08c7bdf1ecb7b69302468ba61d6a1662920f94b97a85efd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8297171e93-tools.jsonl", "bytes": 4313, "sha256": "8df8d48b3622bdc6ecdfb3f6083d7510786c96e9ca6b15b71a737f8ca2ef621c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8297171e93-runtime-log.txt", "bytes": 256, "sha256": "6b9ad6b8b5f2b3291fb65fbc70312c1c31f21e37bf60e9e4346f60b26fcbee63", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/out/customers.json", "bytes": 6738, "sha256": "55741ad10f83e45d112bc72a61142baafe03cf2c5f0ac3603ae66ceb4a073d6e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/scripts/clean_customers.mjs", "bytes": 3133, "sha256": "6bd0bef50eddeb9bd08c7bdf1ecb7b69302468ba61d6a1662920f94b97a85efd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/result.json", "bytes": 1179, "sha256": "c08660c2516a1104b7f815a42b0e9e168175eb37c1901592bffca8a2dc07f746", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/score.json", "bytes": 4222, "sha256": "3d3d24b22737ab63dc4df369078e3330fcc7cd0c0ce8892efa067421a58e56db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/metrics.json", "bytes": 2349, "sha256": "f05e2abe05c208acc7cfb04cfb3b7c166553203c1a40a114872e2f5a9ad81476", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/facts.json", "bytes": 13685, "sha256": "d10a12a95e70f3a0ab6b66e1f004a2100f5ee1d2ca009271c26e18472742b3d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/codex-inline-judge.json", "bytes": 7118, "sha256": "444ee0d274bf92a0409e6423d4fe3beb2fd1ffe6ba047ef71aba0a00deb80187", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-9b-q4-2026-10-22T00-10-11-192Z-eiy2/postmortem.md", "bytes": 11335, "sha256": "a98c91aae150d3129b4f6a16f6a5ca2effdc0950039508548945498bcac12495", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/8297171e93.json", "bytes": 6316, "sha256": "cb61cc6083a4f8aae2a2c7e3792da7ff8d52d40044571768756f8758edec7591", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/out/customers.json", "bytes": 6538, "sha256": "7c4f32dca6bd6cccea77ff9b3db29544d14f7e8053201359e470d937ac5a0238", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2737, "sha256": "b2bb552a4bdbd513df8bf820f8b7dc768b1fe00d44601f29e36188395c284e65", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/tmp/run.js", "bytes": 2171, "sha256": "f87fd27137f54a279d90c5636fa2d0c94a5c46399a05ef646ad2820d6304b084", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/62fd0bcf50-tools.jsonl", "bytes": 3652, "sha256": "f23ca3e0b755cc24e6798783acb64f532d397a704c030b4e2f757a89592c35e9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/62fd0bcf50-runtime-log.txt", "bytes": 274, "sha256": "e9cd04c2bb15a510e96391bfc108349521f2f754cb3a67505632d062b11ebb87", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/out/customers.json", "bytes": 6538, "sha256": "7c4f32dca6bd6cccea77ff9b3db29544d14f7e8053201359e470d937ac5a0238", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2737, "sha256": "b2bb552a4bdbd513df8bf820f8b7dc768b1fe00d44601f29e36188395c284e65", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/tmp/run.js", "bytes": 2171, "sha256": "f87fd27137f54a279d90c5636fa2d0c94a5c46399a05ef646ad2820d6304b084", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/result.json", "bytes": 1268, "sha256": "7615a1b0845adac90056f7ac884bcd15b96fa20700331c1c23892d9705f04be5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/score.json", "bytes": 4381, "sha256": "aaccc997ad08efe4c60920cbdb52f17f50f58fd429cee274d6413fdd5fe79c46", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/metrics.json", "bytes": 1315, "sha256": "42b5c08556bac4fac8344e5202c054f536027ee4d088823dab61ad90e7f16ad3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/facts.json", "bytes": 11506, "sha256": "b42c2147f83a55e85c5a7152d56b9ffde6ed5166e73ec9d481aa44e57b309960", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/codex-inline-judge.json", "bytes": 7931, "sha256": "6bd6e95194e77a5500a12a1523ce2ddc0fea06104f8d414a2f6a928f17e6b9e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-data-wrangle/data-wrangle/data-wrangle-ornith1-5-9b-q4-2026-10-22T00-28-04-119Z-2bf6/postmortem.md", "bytes": 11171, "sha256": "70666abe9367e9bfd2095f8239d2b46d82534f1856b8f862c5116ea337c6f900", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/62fd0bcf50.json", "bytes": 6626, "sha256": "27bc80a456646c9ab3def8800259f5a47c3b508058a0e14ede8856b370c70379", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/out/customers.json", "bytes": 4073, "sha256": "f66d210ee8cfa96f79893cb9680050deb68935115bdc6f493898bce14e74e0e6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3774, "sha256": "61f4eec2f064bd5184b99a7304ecf70a99b942fea223e6dc11b47a7869dd9fb3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/10886300e7-tools.jsonl", "bytes": 3622, "sha256": "002a17ba291f8b43cf5d48efa7029096a494e581b58707d4d9f91cf72a3bf52c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/10886300e7-runtime-log.txt", "bytes": 290, "sha256": "1a3d1736f6f236b907d193d582f72492e9ddf1560b77234292bf30dc09a82a6b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/out/customers.json", "bytes": 4073, "sha256": "f66d210ee8cfa96f79893cb9680050deb68935115bdc6f493898bce14e74e0e6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3774, "sha256": "61f4eec2f064bd5184b99a7304ecf70a99b942fea223e6dc11b47a7869dd9fb3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/result.json", "bytes": 1210, "sha256": "1376629e6de3403ef838ebf538b7a78042c014c803b703209ba441387f424c7f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/score.json", "bytes": 4255, "sha256": "6a234ab5dac97fe72f71e5592809dfc25200b0395f84de6061015a9e66233605", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/metrics.json", "bytes": 2386, "sha256": "433b843db7af6e5a472d9743cd0b19534daad9ca0c992c46a360e8d51c65f4be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/facts.json", "bytes": 13733, "sha256": "ec88561507541392603c68529a7a9ff3571ee0543b2c2d6f95e13c86cb0cf62d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/codex-inline-judge.json", "bytes": 7258, "sha256": "94f40eb07c66cf769819834b2a12e94585d801fca1b0fa8b6c1687d112d0710c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-data-wrangle/data-wrangle/data-wrangle-qwen3-5-4b-q4-2026-10-22T00-57-08-898Z-xghu/postmortem.md", "bytes": 11271, "sha256": "1bbc72c7f2fcaae4aa4325ef52a1206802f3f0f31da67909e2b98312be4cc44a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/10886300e7.json", "bytes": 8761, "sha256": "1af6b98b0cf72d77b5567717cdd915c5d33da567ed1c70ec4ec6174e87f17140", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "970fac708c69a126f7e2dc11a16e8c808d1901b54bfe056f5c45d5d040996c6a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3018, "sha256": "d077ff6677d1647cdb8ca09bb72fc13158d81a4a0d5822aef1ca8d5ec534d1c1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/5e24b4a2f0-tools.jsonl", "bytes": 2497, "sha256": "46a88afbb0bbdc8c9a6dc4df7f9390afe038f7ac2e751a37116ec4c3ac404f9a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/5e24b4a2f0-runtime-log.txt", "bytes": 265, "sha256": "cc5a019043d08aaa25482fd028942f1192d6e2bd76411442ce7b4f7f76f196aa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/out/customers.json", "bytes": 6986, "sha256": "970fac708c69a126f7e2dc11a16e8c808d1901b54bfe056f5c45d5d040996c6a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3018, "sha256": "d077ff6677d1647cdb8ca09bb72fc13158d81a4a0d5822aef1ca8d5ec534d1c1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/result.json", "bytes": 1318, "sha256": "3adbf0813306291d23058be5dfb43ca060a0004dcab6cfeddf925a92323dcb81", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/score.json", "bytes": 4434, "sha256": "9490048d72e3cbb1785d799c6bf8d8ba944515031852db2e2feca2e3055a2aea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/metrics.json", "bytes": 2414, "sha256": "4fee4274e98936deaa98ac429e83ea2e7c08edabafcd4840a97684d43517f844", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/facts.json", "bytes": 13108, "sha256": "bd870e63a8f1e0e019a8af8ec1c226883cac959c897dc4c82c34ac0a45656b33", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/codex-inline-judge.json", "bytes": 7477, "sha256": "11c3db0a6054cb6fa9b9b9c55078e404877c0a0063c865ba18b84c0f5cf37e91", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-data-wrangle/data-wrangle/data-wrangle-underdog-saluki-27b-iq2-mix-2026-10-22T01-02-49-558Z-no4v/postmortem.md", "bytes": 12679, "sha256": "5e374edbefc3ecd9109c5a226a73e8fe8e5842db84351e28eeb3184d21d9501b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/5e24b4a2f0.json", "bytes": 6615, "sha256": "9ffb332524a1cffd93c36399dc2380d60d0fad3ddf258b6d0c86b2b7c4ba1107", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/out/customers.json", "bytes": 5978, "sha256": "328c66b6c8b2c72ee05a6963caa071ece7dd43939e0f75ff69367cb62fc4c902", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3350, "sha256": "6d017f0ac66f1b593f512485a0d39c83c7e6dd642ea2a4e19e02f3fc35ec8350", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/b5aebdf59b-tools.jsonl", "bytes": 1848, "sha256": "f2ba11e58f58a2b2d9556ab64a6f87266111418e27835d64b7c89ebb8bcf24bc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/b5aebdf59b-runtime-log.txt", "bytes": 274, "sha256": "e6bfab589a98567e7379a8945e6a2563eae3b4d342bdc3a74b85816e503351dd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/out/customers.json", "bytes": 5978, "sha256": "328c66b6c8b2c72ee05a6963caa071ece7dd43939e0f75ff69367cb62fc4c902", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/scripts/clean_data.mjs", "bytes": 3350, "sha256": "6d017f0ac66f1b593f512485a0d39c83c7e6dd642ea2a4e19e02f3fc35ec8350", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/result.json", "bytes": 1265, "sha256": "ca250d77fc2e55a790afed3c1b6617ed4c854f5b92f2e7b389da9810d22442a1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/score.json", "bytes": 4406, "sha256": "af9d2834628992bf7d029f4a08a2ca0e7bbf0c207278269f720fee217acf8ab9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/metrics.json", "bytes": 2340, "sha256": "2897d35ae2a34220f3191c118aa79918cc261fd5e940c4b0e9c798f3700b5043", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/facts.json", "bytes": 12925, "sha256": "e8b3b3bfeaca4c2e4b00b590119a12fb9f350c3cd474f358a48d0f1c62b9e455", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/codex-inline-judge.json", "bytes": 7098, "sha256": "a67f2db1efebedf265da41556a011eb35ab449685862b4f84aeda19b5be7a9b2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-data-wrangle/data-wrangle/data-wrangle-qwen3-8-27b-q2-2026-10-22T01-48-22-161Z-3c9j/postmortem.md", "bytes": 13644, "sha256": "777f8ebbfb8d298cc11221b5a9ce80aee0c727d48c47613201f374c1f95c762a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/b5aebdf59b.json", "bytes": 5930, "sha256": "64f50ddeb341591f89ae990f9740ce345e2d48c6a1f784b93bab095743ebaf3b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/out/customers.json", "bytes": 9107, "sha256": "8fe0c4ac5f504fd8d206b6e18967127e1205baa6492fa859b63e3e036aa3a026", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/scripts/derive.mjs", "bytes": 3784, "sha256": "e707de6757b9892c7347f34744083bd1301f73ab7875a712ca4a1d04590ed552", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/42e133d613-tools.jsonl", "bytes": 2228, "sha256": "01fb1482ff42d8fa9dccd68df57245c0c34d1a362c9988976eb39b55038bbeb3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/42e133d613-runtime-log.txt", "bytes": 255, "sha256": "b003a728a1e5308f1c071ba69ad9b7ce9d13229bbb662f154fdc97de1206c348", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/out/customers.json", "bytes": 9107, "sha256": "8fe0c4ac5f504fd8d206b6e18967127e1205baa6492fa859b63e3e036aa3a026", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/scripts/derive.mjs", "bytes": 3784, "sha256": "e707de6757b9892c7347f34744083bd1301f73ab7875a712ca4a1d04590ed552", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/result.json", "bytes": 1207, "sha256": "d33bf236bfcd702ad5cea1628725a970964626b7bce93b3f87a89a19752c80e4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/score.json", "bytes": 4061, "sha256": "26de20f5ab5905018f08742aefd3d29bf51401fc92a38a56e73d949f48e1d31b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/metrics.json", "bytes": 2357, "sha256": "d03519ebc1e4bfb086d38e7b96f4ec67466c848838e68d3b96d623eeb73b078a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/facts.json", "bytes": 13837, "sha256": "c55c2e561b9032763ce1f1a0d35d8485c0ef8ed8b14186236f93cf97c62043e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/codex-inline-judge.json", "bytes": 6822, "sha256": "ebe2cd3f2a8cded596202d5ab129d4c44620803aba3a825a4f189d8d5f4bca4f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-data-wrangle/data-wrangle/data-wrangle-gemma4-12b-q4-2026-10-22T02-39-11-991Z-nim4/postmortem.md", "bytes": 11899, "sha256": "13d9f6df15da38ab1ad46f63c2a1677bf35fb7a2ecf5a360ca053932c64eea2d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-data-evidence/42e133d613.json", "bytes": 8733, "sha256": "8c9ece2c2f08c2795ec25810e4b3933bc235bbaacee2cf6a417a3b1c7e08ff81", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 7583, "sha256": "040d5fd6fed0b96e08718cff6344fdb3a4a33ea534fd4235e40585f65502e1c4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 7583, "sha256": "040d5fd6fed0b96e08718cff6344fdb3a4a33ea534fd4235e40585f65502e1c4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/result.json", "bytes": 1017, "sha256": "a2fcdd7b87da924f16be079de73745e5d928c53d629ede70eb4e2eff5924a2eb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/score.json", "bytes": 3039, "sha256": "280b1b219f3f5a500d16e4d1ed9d10e4df51fa8828d8a3a34cfd45a419b81e47", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/metrics.json", "bytes": 2368, "sha256": "a5b148427c0e00fe95963d06d1bfd2333b0656e6cfd226566bb3e303a6d776b4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/facts.json", "bytes": 13415, "sha256": "7be66a25c1cc90e729eae0a76b8647551e225bcadd29ba29ca10c94ef9d760e1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/codex-inline-judge.json", "bytes": 6185, "sha256": "8c2133b9373db1c46c883a9e67722eca2fcce01bd8132d21dc773a51a830f77f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-9b-q4-2026-10-22T02-52-57-314Z-wilx/postmortem.md", "bytes": 9947, "sha256": "88d975dc9489aacabafb35f21bfe14143b048fa1d5866d4b8064dfb411ec223c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/1a66f886a3.json", "bytes": 6060, "sha256": "710fe4581e63d014b809edb4c3bdee3da09816fbfde601925a91ff277578ddd4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9304, "sha256": "ba4598b8ecb3835912bf6182ac760a75416e8315dae8490792010e1329fe460c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7e4feafcef-tools.jsonl", "bytes": 864, "sha256": "a62a7be0097ca8f375270330fecc5ca29127ee0d268a3ed159f7b2021aed18f0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7e4feafcef-runtime-log.txt", "bytes": 324, "sha256": "c3cd7689534357dc7139c6a45abcd430c1b3f5bbe9c3475ed7bf108223799d50", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7e4feafcef-gate-excerpt.txt", "bytes": 334, "sha256": "c0103baeea4055bc1a70f74b90f67ceb263893e3aedc66327dd93a36d6bfe1c8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9304, "sha256": "ba4598b8ecb3835912bf6182ac760a75416e8315dae8490792010e1329fe460c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/result.json", "bytes": 1326, "sha256": "5cfe9def875298226b63315b27a17fe29b3f62afde9c2b7f864d3ce3b6c37f00", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/score.json", "bytes": 4546, "sha256": "057a424357d70d4b9b3ef7d895540a2c468ca192970a8aae309dbd0e51f78994", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/metrics.json", "bytes": 2383, "sha256": "49c7ce13c887b8ffd6334051fd494dc8841cb02839d79c82d6b15ad225271469", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/facts.json", "bytes": 14717, "sha256": "057b4d0c1161994ea80fcc1186397c852c97117fe9cd71f5b573f5557f4fc110", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/codex-inline-judge.json", "bytes": 7782, "sha256": "c7126fee690fecfa4ada0aaeef1a8164596e3b54ccd308a80a17834226d71940", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-ornith1-5-9b-q4-2026-10-22T02-55-56-156Z-h4y7/postmortem.md", "bytes": 12734, "sha256": "003a72ad6355f4d5785cc120fc1c14df9bbda7dd2ac3182a6a027f2299332a00", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7e4feafcef.json", "bytes": 6114, "sha256": "744b5fcd37d63fd71f63e302e76b0550bc6b83615b8cb51f06b33f649146d6ef", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8052, "sha256": "5c6d1cde53a01d8895f1de190da887afbbc79e2c0dc7a42180f9f9a034252692", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8052, "sha256": "5c6d1cde53a01d8895f1de190da887afbbc79e2c0dc7a42180f9f9a034252692", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/result.json", "bytes": 997, "sha256": "59d6426ff6d757297a47ba238f7f141eb6c819c085f6fe4de97f17a6ac99793a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/score.json", "bytes": 3052, "sha256": "39167d9cacc370e8e2660c1e63e6a192b62cf9f8756207e573f5bf83280e8f4b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/metrics.json", "bytes": 2379, "sha256": "060b8dad97da41d3be0bec60e9bc8aa8916bc22dcb1f650eff4684800ac5d252", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/facts.json", "bytes": 12976, "sha256": "d5c1e276244089ae6505747514e5bd9470285960eee4bf9ad9f4a34bdcf17566", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/codex-inline-judge.json", "bytes": 6210, "sha256": "fe28d01f6968239028f80db936ca7729abebd208dfdf841af5d7c045ef81133a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-5-4b-q4-2026-10-22T03-01-07-863Z-9xwh/postmortem.md", "bytes": 10118, "sha256": "0e9739ac2c34e23915c772471593a768ee6b6b4e19e6433756f5c384f255cf47", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7dd90a1563.json", "bytes": 6060, "sha256": "2b9ba3113d212faae829450c7309c3528b37eebadc369af74f39a206c389eb75", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 10270, "sha256": "39fbf473ec404ef6aff8c71502f22bde10c5a9519437feddef425ec585c22b35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 10270, "sha256": "39fbf473ec404ef6aff8c71502f22bde10c5a9519437feddef425ec585c22b35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/result.json", "bytes": 1089, "sha256": "024a10b70a1fae6ddd205aa1a5073d437a048ec7db7a926aaafe6339c00f5050", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/score.json", "bytes": 3081, "sha256": "382d17826475dfaa19e22368fc9d9de5d75f54af0a5208597818d29fdada1926", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/metrics.json", "bytes": 2410, "sha256": "9a045956d393244168dd2b45a2c3ebd1a83412205cc0fd5db423ae8a0023cad9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/facts.json", "bytes": 13618, "sha256": "e8939f48ef04f3f84191c22b15c47ce0b79b3159e6357118ef886e309995f935", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/codex-inline-judge.json", "bytes": 6579, "sha256": "c3c18fc50b7406d4243b6e1728f2470714bd3cc1f050d462f3b3cb801693e45d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-incident-postmortem/incident-postmortem/incident-postmortem-underdog-saluki-27b-iq2-mix-2026-10-22T03-04-11-127Z-nce3/postmortem.md", "bytes": 10018, "sha256": "11236d14f12d18331b1251cd44f39a7b826dc801eb4a45398bbd3859b8c1d70e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/836230ae45.json", "bytes": 6440, "sha256": "599abb0ef0b62b8a5bf7e18d61e6825a0f1db720456e805ea7921bc7f68d59d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9535, "sha256": "0aca991d68da4f21fc462b06868dd7382d49fcdfda28b5a147577550d3a3fade", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 9535, "sha256": "0aca991d68da4f21fc462b06868dd7382d49fcdfda28b5a147577550d3a3fade", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/result.json", "bytes": 1024, "sha256": "dbbb69289956a72c1f0ce52a1ce3154aebbfd2707f3318a03d0177dd5e7647a5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/score.json", "bytes": 3054, "sha256": "fe6e6aa73fcbcaac235bfa34cd4f27ccb70b68b808dc3bf7b1a097be88fc751f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/metrics.json", "bytes": 1316, "sha256": "255a40629c2a8b2277ecf3ada66d55e333c8683ef38d3479724e45e172ac3127", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/facts.json", "bytes": 10923, "sha256": "42879205acb4c325619813aca478a2cfeab80f932806adcbc5be09c8eb4c0396", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/codex-inline-judge.json", "bytes": 6228, "sha256": "c787c98dd49fae222ad9397d88bf0bb0968396d0815f8d04b993c3bee49669ec", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-incident-postmortem/incident-postmortem/incident-postmortem-qwen3-8-27b-q2-2026-10-22T03-30-31-715Z-pjqd/postmortem.md", "bytes": 9704, "sha256": "d80f129ae4da6eed8c2b4fe4c4b7bf4468c962729f00703aa78c99868507b4a5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/739d1f5b43.json", "bytes": 6087, "sha256": "99e54c48d81ec8cc4e53dc3c1b53f9134e12a3d8f199ca9225eda9572b94c6df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8169, "sha256": "d035e3b46247c88a81d4d86de494a015cca0366698a056f458e2ed7ae7ea9cdc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 8169, "sha256": "d035e3b46247c88a81d4d86de494a015cca0366698a056f458e2ed7ae7ea9cdc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/result.json", "bytes": 1018, "sha256": "255ef22bbe757854f7c8581cd630cc003e24d3287078739d27be4018e1392913", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/score.json", "bytes": 3055, "sha256": "f34d1512910f0c703a39a4ea2ac91f261794402ce102881473de9ece424f605a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/metrics.json", "bytes": 2384, "sha256": "c2f5cbfcbb147d77d6996f5e9a7bfd1d31ee9cb39f274b2e38b9b4b67492fc8c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/facts.json", "bytes": 14348, "sha256": "b84f8ee5320ea3a825cd5b55fd473f77d9fd3f762802a2f1f53f90142c408014", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/codex-inline-judge.json", "bytes": 6218, "sha256": "37d7b03b995b6c912fb1f97c1f018dff0ddb2f84718eee6a70e1798d856d2bde", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-12b-q4-2026-10-22T04-04-46-948Z-yr7j/postmortem.md", "bytes": 9672, "sha256": "66e9513c4799cca628a673ade9215c174caabd74c2bbfbf1d33f3da5a5992f87", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5dd5a06512.json", "bytes": 6060, "sha256": "f105a2748ad0991d244cf14b2fe2ac471827e84a5526e8fcbd7b8e46e12b5d57", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 7838, "sha256": "618d1b20da59df3dcc872e88140d66616ce103d7c438f9f5875b5b494ba0daf7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/postmortem.md", "bytes": 7838, "sha256": "618d1b20da59df3dcc872e88140d66616ce103d7c438f9f5875b5b494ba0daf7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/result.json", "bytes": 1017, "sha256": "252b72617277f5cf44a99174a071ea6f5fb2abaf54ebb700a00282b51ffee32f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/score.json", "bytes": 3043, "sha256": "6ad37d2c95373189d4febeb6e138ee3aeabc6fef77a7e5b3413952c823ea765b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/metrics.json", "bytes": 2377, "sha256": "6e04643c4fbe7b514561bcf746e7f1e3b588d484c58026be2f91fb13f52b8154", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/facts.json", "bytes": 13316, "sha256": "5604a178c1fbe3eb66f28b2e110336aa5260a62daba6040b576629af41b3d709", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/codex-inline-judge.json", "bytes": 6202, "sha256": "61e85b554dc4b0f1a3beff1c72727c0a788e600328eea59e016876cbbbed771d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-gemma4-e4b-q4-2026-10-22T04-09-06-062Z-7qu2/postmortem.md", "bytes": 9574, "sha256": "786b3001b4e3ed12d0eb8fe33882d59b15e4d723b53911d1ffcabaf29fd3376e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/28624a466f.json", "bytes": 6060, "sha256": "825e0969b2268316cff72ec7382798f1e649c5e2bbbd252a0831c2d92b3f18d1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/runlog.md", "bytes": 884, "sha256": "87ae300ae5c0466a889534196a6a59976878b4ad356f495f6f19c5b2b7cf48d8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1261, "sha256": "8abcc909e519b7232f4539db304b89c123d73ac38e0defd13db0af6cc9082411", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6f1ce3c25c-tools.jsonl", "bytes": 2218, "sha256": "926a19acca420a15ea2f99819e9336f3cbbf47014e7bcfa929cb706d4cfe949f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/runlog.md", "bytes": 884, "sha256": "87ae300ae5c0466a889534196a6a59976878b4ad356f495f6f19c5b2b7cf48d8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1261, "sha256": "8abcc909e519b7232f4539db304b89c123d73ac38e0defd13db0af6cc9082411", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6f1ce3c25c-tools.jsonl", "bytes": 2218, "sha256": "926a19acca420a15ea2f99819e9336f3cbbf47014e7bcfa929cb706d4cfe949f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/result.json", "bytes": 1005, "sha256": "21e493e2d6ace763823b1798e3458f2860e24512123e5b3766340498ab863e76", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/score.json", "bytes": 2988, "sha256": "73cc2630d3e8bc682807a6a80a27f3da44d2f1834506b6358967a769695515c1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/metrics.json", "bytes": 2370, "sha256": "93841e070093835c0942d13a501589b48741ea7583d9a99821072352adbfcf29", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/facts.json", "bytes": 14684, "sha256": "5ff7163cd31e4eef3c07066cfbd661569ca0d5726e5d69e06ec070b5765f1f88", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/codex-inline-judge.json", "bytes": 8606, "sha256": "0c4b398803770c63247b07748074b167cacaa032ef7a335df0e8b5e783cc8ae0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-ornith1-5-9b-q4-2026-10-22T04-11-45-793Z-s0ej/postmortem.md", "bytes": 9662, "sha256": "6d607172c7a8eb3e2c4c33f5f3bf417a8e3c9bb6c62b51512460ed485e7b8b11", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6f1ce3c25c.json", "bytes": 12278, "sha256": "f69001ae7c55198a11092bf326485de633205abfb7eb33d4a97891c55cc0221e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/runlog.md", "bytes": 720, "sha256": "b021442b25372f2b45df808e86b8c0f7f9d82ec7013d79a1b8f5697a89d43026", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 457, "sha256": "390c2255f8e619973f4a1a34727c758c1e6ff62f25977f4eb40e77b2a5ea407a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6c2d7e5b54-tools.jsonl", "bytes": 2845, "sha256": "2075e57176a803f38b756a9464e0328f34736d47c5e22d056059eb05e2b69574", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/runlog.md", "bytes": 720, "sha256": "b021442b25372f2b45df808e86b8c0f7f9d82ec7013d79a1b8f5697a89d43026", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 457, "sha256": "390c2255f8e619973f4a1a34727c758c1e6ff62f25977f4eb40e77b2a5ea407a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6c2d7e5b54-tools.jsonl", "bytes": 2845, "sha256": "2075e57176a803f38b756a9464e0328f34736d47c5e22d056059eb05e2b69574", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/result.json", "bytes": 993, "sha256": "c99bbdb1d45cde24a6b5b88780d058bb6b95b05a50a33270dec8608ccd7ac52a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/score.json", "bytes": 2998, "sha256": "2f996b442fae5fb398c3197e77e834fd9f4ed39cad6eb4af30c185f03f0503d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/metrics.json", "bytes": 2339, "sha256": "1e7cf25e98353977d43ccd386886103d1cea5302c8691a1f2aca7d8c55fc5ff4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/facts.json", "bytes": 14729, "sha256": "e7db14d4a7e3cd279c06af0bdd1d3c651745d48068ef67a992d1f29a432991b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/codex-inline-judge.json", "bytes": 8518, "sha256": "445becd926d5b1b938091ce136cf3048bdfb6efceb0529f058505282034388a7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-4b-q4-2026-10-22T04-14-27-349Z-4z8g/postmortem.md", "bytes": 9657, "sha256": "7a9dd39ca8bbe946235420a742255b4438ca7545a04e4583263103cfc5fa5023", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/6c2d7e5b54.json", "bytes": 13064, "sha256": "075805d2f2c7b2bcd2ccb80c8138bdbbe501ea85f717b4c3dfbc2779099d8543", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/runlog.md", "bytes": 137, "sha256": "715909b4cb3d0f96488dee6896f4025f9334b72455ec9d17b36b9850b6bdaa3d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 542, "sha256": "78140de4a57e7c2d1818f1901034bcaa37d012dec34ccb9dd94c0daf4b1cf94f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2a6bae2ce6-tools.jsonl", "bytes": 1678, "sha256": "725af0ba0443d4d86d31d324aa7858694c4b3b87792aee76a24322d2500b3198", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/runlog.md", "bytes": 137, "sha256": "715909b4cb3d0f96488dee6896f4025f9334b72455ec9d17b36b9850b6bdaa3d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 542, "sha256": "78140de4a57e7c2d1818f1901034bcaa37d012dec34ccb9dd94c0daf4b1cf94f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2a6bae2ce6-tools.jsonl", "bytes": 1678, "sha256": "725af0ba0443d4d86d31d324aa7858694c4b3b87792aee76a24322d2500b3198", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/result.json", "bytes": 1066, "sha256": "54cc58aacd3a5fc77412b840e36345cb4ba6e011a38a2d51e2ea1427c88882c2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/score.json", "bytes": 3009, "sha256": "d3a1d90b03ba5ab5261fc4f1e6ce239496321d4992efef126cbd59a894d2f891", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/metrics.json", "bytes": 2363, "sha256": "63715a2224b66ba6e15ee472b8cabb7c0edde83cdadd7f34b800db81d8660b4a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/facts.json", "bytes": 14112, "sha256": "cef848b57a5cb4ccb00f560435b55c3c7fbb601cb2fe7b89c8f55b910df92fa6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/codex-inline-judge.json", "bytes": 9045, "sha256": "8b4c1e8c6d19cdde9bb8c3a81a3deb12f2eab2b037fcd7bb307c84234c15a5bc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-underdog-saluki-27b-iq2-mix-2026-10-22T04-16-55-700Z-xwsy/postmortem.md", "bytes": 9571, "sha256": "9ba6a908a7f12895a379d194ec65f072d36e713d6d0a45c9906ec4cae1132ccb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2a6bae2ce6.json", "bytes": 12077, "sha256": "ed1268d04b0b74ec5dfa7d88f107c8fba7c205f95b2054bf8fb6ab93f3880c68", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/runlog.md", "bytes": 126, "sha256": "dda03c20cbe36f1736e0d0f8f1edc92569c8eb292b5c1bc02d878138bcd77f2f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 655, "sha256": "e712232178aa01050b675565ee782a6f296596c05d27584889ef42fe0280a88b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d8d474133f-tools.jsonl", "bytes": 1698, "sha256": "f9245b884fc259c8f9bf163790e6c95d7834bcfc91c396b7f9189599ac2936ea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/runlog.md", "bytes": 126, "sha256": "dda03c20cbe36f1736e0d0f8f1edc92569c8eb292b5c1bc02d878138bcd77f2f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 655, "sha256": "e712232178aa01050b675565ee782a6f296596c05d27584889ef42fe0280a88b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d8d474133f-tools.jsonl", "bytes": 1698, "sha256": "f9245b884fc259c8f9bf163790e6c95d7834bcfc91c396b7f9189599ac2936ea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/result.json", "bytes": 1001, "sha256": "a02afa335f6f950bbe6debd303815886883ebc9a8d4b667418a400e09512f035", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/score.json", "bytes": 2986, "sha256": "524a6b787a929f05102a017f2acd7617ca536ff4d2b3c8e3138281813e5843de", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/metrics.json", "bytes": 2331, "sha256": "26985f71a51cfa45cbd4c5ae12031da8f9f922cfbb948a976a958dad2a54d450", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/facts.json", "bytes": 13980, "sha256": "7b964299909fea2e5023a7e5f47e4de739cdaf8263f9cdfaaabf10c01fc06b55", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/codex-inline-judge.json", "bytes": 8537, "sha256": "c7a9ef9fe409bb1fe68fad137b25b4a0f47963abcdb83bf84101bcf5840998f3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-8-27b-q2-2026-10-22T04-21-22-502Z-fw2o/postmortem.md", "bytes": 9485, "sha256": "a2425e44e0d00be5a539e9075de5afefcb3be35b4b18e72f9f14e530fceb277d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/d8d474133f.json", "bytes": 11512, "sha256": "6cf3ebd513070753c7a18e5856baaa0083cf621f94800e3a37f075166b244fac", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/runlog.md", "bytes": 122, "sha256": "dd7aa4ce1f8537522539832ae88fc64f8e1e3c2dae6e73e6fb591c8193b115e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 390, "sha256": "f6b6666c5cccb180cf622e86934962b125c89ef6e37c698d3d041308b90e2c5c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/0d1c8bbd69-tools.jsonl", "bytes": 2867, "sha256": "acad6aaa9bdd0a2440ac3ffbc90023e86ea579690efadedb399eedd4dc6400f7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/runlog.md", "bytes": 122, "sha256": "dd7aa4ce1f8537522539832ae88fc64f8e1e3c2dae6e73e6fb591c8193b115e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 390, "sha256": "f6b6666c5cccb180cf622e86934962b125c89ef6e37c698d3d041308b90e2c5c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/0d1c8bbd69-tools.jsonl", "bytes": 2867, "sha256": "acad6aaa9bdd0a2440ac3ffbc90023e86ea579690efadedb399eedd4dc6400f7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/result.json", "bytes": 996, "sha256": "181c03a8f3e703febe594df6a96816584b55a15b385f31167d67018988e1edc8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/score.json", "bytes": 2998, "sha256": "5192a67ac02b65b638e975e00bd90288104e97f11237716d34f7dbb440e23a1a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/metrics.json", "bytes": 2372, "sha256": "638d2e050af6d2cc3a0f42cb86a2dd42c17418bf3a157372ee4860aa505973b3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/facts.json", "bytes": 14619, "sha256": "0be0bf9f6932121ea9e41ec2daa1e049e5c17fb34a2a2bec1855afb0cca4f0db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/codex-inline-judge.json", "bytes": 8501, "sha256": "0ec753f475d049a9603e959500f5ef23be12be2837352fc6cee2cbfdcdae9e73", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-12b-q4-2026-10-22T04-28-07-686Z-4gz4/postmortem.md", "bytes": 9639, "sha256": "ef18c9befd1f0b0d68e82f087fc882aed7228a8cb38a913bc7833dc792654b50", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/0d1c8bbd69.json", "bytes": 13086, "sha256": "2f5784216560b1f2e9cdede7ae8e54de31d87b6f66bb77405e61f120f2133aa9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/runlog.md", "bytes": 309, "sha256": "498a2605167737c1043a392b48ab52e8a017b854d46987e94e4fb175d4fc04a7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 420, "sha256": "06fdfd3cff0728fcc96950aeb9ea3922bca8f9d56f76c26a6d8a4d19bd624e44", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5cefa7cb82-tools.jsonl", "bytes": 1961, "sha256": "d92346b09ae47b38e7d8b37e11b250886544635e37a3f9e75f2d458ace4d5fe8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/runlog.md", "bytes": 309, "sha256": "498a2605167737c1043a392b48ab52e8a017b854d46987e94e4fb175d4fc04a7", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 420, "sha256": "06fdfd3cff0728fcc96950aeb9ea3922bca8f9d56f76c26a6d8a4d19bd624e44", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5cefa7cb82-tools.jsonl", "bytes": 1961, "sha256": "d92346b09ae47b38e7d8b37e11b250886544635e37a3f9e75f2d458ace4d5fe8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/result.json", "bytes": 994, "sha256": "8243589fda49464badd347535ba9e54a3cd9866410e3caa815042124d654802f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/score.json", "bytes": 2982, "sha256": "c08c92e27fa7c81feb0a8076c1358fce59757250cbc0fc81053340008a6c0e99", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/metrics.json", "bytes": 2331, "sha256": "ad115729fee86283604b1428888a2eb9c0a9ae2308c7056426ae17da3f3c017f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/facts.json", "bytes": 14631, "sha256": "a48c2930bb9a0a3e517740ad8b1ae03be7d89ca3b7423a948e3f910e8df2ad9e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/codex-inline-judge.json", "bytes": 8491, "sha256": "1c63081c03dd5aa586970449c0b6dbf4a8fca8caceed788d8371151a57d98b48", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-gemma4-e4b-q4-2026-10-22T04-31-29-070Z-5gws/postmortem.md", "bytes": 9491, "sha256": "eeb48186df65b6369ee61e4b2984ae13cd017104a1c82932179212b6ae1a6f81", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5cefa7cb82.json", "bytes": 11830, "sha256": "6043aa8f5c273cbfb54945b5a59614d6885dc0d9d306a8f60d5335aa2e056ff1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/runlog.md", "bytes": 126, "sha256": "dda03c20cbe36f1736e0d0f8f1edc92569c8eb292b5c1bc02d878138bcd77f2f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 309, "sha256": "2975c23a098a7f33a4477a2995e93d7917e2a0c5df7f8ca1e9e1234424e226e8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b531fdf7c3-tools.jsonl", "bytes": 1698, "sha256": "cdabbb5161ce1cedc8518ad166fba6933afe413ac7db3bf9d0ac18cb1826ad43", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/runlog.md", "bytes": 126, "sha256": "dda03c20cbe36f1736e0d0f8f1edc92569c8eb292b5c1bc02d878138bcd77f2f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 309, "sha256": "2975c23a098a7f33a4477a2995e93d7917e2a0c5df7f8ca1e9e1234424e226e8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b531fdf7c3-tools.jsonl", "bytes": 1698, "sha256": "cdabbb5161ce1cedc8518ad166fba6933afe413ac7db3bf9d0ac18cb1826ad43", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/result.json", "bytes": 994, "sha256": "5941c10aa43e19cf4956045c17fea94b986f30fb6a148f299a16940c4b4fc9cb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/score.json", "bytes": 2982, "sha256": "b5fec2835decaa4d7b96e6b9257525cdd73a934ebb2fadcb37b9c7c9146a5ad1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/metrics.json", "bytes": 2323, "sha256": "8f94f24bb396d25dd5e3946679e60362f5dc9b0385bb1b2279c6167d91e12752", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/facts.json", "bytes": 13474, "sha256": "91f7cc5ede10bdbec9b953df757a004dd78a9adc56818984140af3239034e2f6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/codex-inline-judge.json", "bytes": 8501, "sha256": "5604728deb036fe3e4d244f10445ab7cca9961cbe3963d3a472f468474293606", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-qwen3-5-9b-q4-2026-10-22T04-33-34-810Z-5u8y/postmortem.md", "bytes": 9498, "sha256": "bef38147ba055e365df1678891aaca7ee4c0963f8f0d531f7edef576e69daaf7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b531fdf7c3.json", "bytes": 11467, "sha256": "2ef52c5bed19b29b37e6999609d7ffed60646c54b900cd3baf72c18e17ed05f8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/workspace/harbourview-office-relocation/plan.md", "bytes": 4095, "sha256": "e6b03218764e4e873c1e333a9b5f0a059f167c2302f209f58e6db785ebe266bf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/workspace/harbourview-office-relocation/plan.md", "bytes": 4095, "sha256": "e6b03218764e4e873c1e333a9b5f0a059f167c2302f209f58e6db785ebe266bf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/result.json", "bytes": 902, "sha256": "639a3aafa8592fe98d7efa088298b68ef01be64d53c6f9a02c1698f7767b7b59", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/score.json", "bytes": 2891, "sha256": "35932475b5dba1541122c55dbbc2aad5f7892028ab05d61bb19c64bc94f33bf6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/metrics.json", "bytes": 2334, "sha256": "f88139f14134ccd0d629016ac80c8e5d4f017b566c0a96396bfc1271914c0439", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/facts.json", "bytes": 11940, "sha256": "013f210d2832b9adbd3e693f0814680039ec1032516686764b246a7b618dd37b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/codex-inline-judge.json", "bytes": 4066, "sha256": "08590515e4ce307ac0ce1bb2613a74fe94483bf63ed871a613b3dc1eba1dfdc7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-4b-q4-2026-10-22T04-35-37-032Z-sym8/postmortem.md", "bytes": 9734, "sha256": "fb29462e2adaff49b7ccfd82f47b2fedcfbe797690ee8f149792159deee9f05e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/c0534ce639.json", "bytes": 3658, "sha256": "8c777391e53f685f8e136327b15d09268eb2f7fa23d74221eaa3838e5241ff45", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/workspace/harbourview-office-relocation/plan.md", "bytes": 3483, "sha256": "dceb6f674835147f8784bd20ee770e4668535849dbf8f3ba9a464bc825651104", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/workspace/harbourview-office-relocation/plan.md", "bytes": 3483, "sha256": "dceb6f674835147f8784bd20ee770e4668535849dbf8f3ba9a464bc825651104", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/result.json", "bytes": 975, "sha256": "b07b9c7160d0f821b0d6145eb3c3a7d5fc7fd129f1ce72b19562e58ac91acc08", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/score.json", "bytes": 2933, "sha256": "916fb005293bf1e2b2106def280a726a5fc8baefca1adb22ffcc020f5a6445be", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/metrics.json", "bytes": 1311, "sha256": "39e26357454f3f53508a46b4f44bfa4e304d5a5ca205483864e709b2935d62f1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/facts.json", "bytes": 10378, "sha256": "296bcc22e776010b1f00a0a8b433588d86bb2fcb4a2cb3bc6fbfad933c7a1288", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/codex-inline-judge.json", "bytes": 4308, "sha256": "cdb06fc4b761968c5250115895efd62cfffc372ddcf9b47559406263dc2dd1db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-underdog-saluki-27b-iq2-mix-2026-10-22T04-37-33-109Z-dorc/postmortem.md", "bytes": 9729, "sha256": "107621f1d58cf1db522785db43b541cae6bd6892ed0b75f56eaccd231a50e723", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/bb5cc90e34.json", "bytes": 3896, "sha256": "23b2e7442e9863c13e588e208eedef5221ecf6baeb2c5bde9a62f6829d9f8dfd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/workspace/harbourview-office-relocation/plan.md", "bytes": 4922, "sha256": "33d5125ec7d81a4d9cb06aeedffc3e1d3938e992f36c849f5a69c38611050c25", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/workspace/harbourview-office-relocation/plan.md", "bytes": 4922, "sha256": "33d5125ec7d81a4d9cb06aeedffc3e1d3938e992f36c849f5a69c38611050c25", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/result.json", "bytes": 910, "sha256": "596f66729dc8472882f649061e880b181bbd9203df6905c374f5b066ec2638a1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/score.json", "bytes": 2907, "sha256": "2575d9b80928d2d4e525dd3db01c2cfb65388ba9f273876f9266e04d48748f78", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/metrics.json", "bytes": 1309, "sha256": "cf0161199fcba6fdee3ca09769990e6bfbb023579d87b0e7804cea48313f97c1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/facts.json", "bytes": 10230, "sha256": "ef20d684f88216cec9080d04985a0cd6971c2a8daa2e34c9bcc5d634e99f6445", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/codex-inline-judge.json", "bytes": 4084, "sha256": "6c6e352f5b8935f388a84b20fc2db746155bc7a38a2d937e69e873a790b65a99", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-8-27b-q2-2026-10-22T04-40-58-673Z-4t14/postmortem.md", "bytes": 9662, "sha256": "67d42415ed764b81f818b61ff590d094a3e55d3e5ff5ffbb1dced9e948ff9e5e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/68a57cd9ee.json", "bytes": 3675, "sha256": "d890a2df79934f999011af2b9de6f28b873dd4e044db84078f3f625acb57bd22", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/workspace/harbourview-office-relocation/plan.md", "bytes": 2117, "sha256": "a64f030a5d48c0b44c3d0e8de891696684cf279d1d318957dce31145fc176478", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/workspace/harbourview-office-relocation/plan.md", "bytes": 2117, "sha256": "a64f030a5d48c0b44c3d0e8de891696684cf279d1d318957dce31145fc176478", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/result.json", "bytes": 904, "sha256": "b4c91778f3f34070755761dcce083fe6afc683607adc624def0dcc56e66c4dfb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/score.json", "bytes": 2891, "sha256": "6387d165e2b8c7893973e0141bdcbfb21cc359bcf8bfe883f1f1465724b01e26", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/metrics.json", "bytes": 2363, "sha256": "002b220427b19ddb2a17b1be681899cf9ebd24dfbb6171cbdb2c1df267e40c82", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/facts.json", "bytes": 11349, "sha256": "7f021e548b229ddc5af3cd00a146be1573e700981127bc21f86ca797617d2240", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/codex-inline-judge.json", "bytes": 4063, "sha256": "a1320bcc8c81fdfc5190427e3a608b69efd6b2db81e2441b63215f1971be2a82", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-12b-q4-2026-10-22T04-48-51-064Z-sdny/postmortem.md", "bytes": 9546, "sha256": "3c461edf06792739213c5ced05f9fad0934bf0c48d8b4e7986535a5a29a17b94", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2051814134.json", "bytes": 3658, "sha256": "c0cd43be185c3193ba26a4ffccfce65b1e4df0f94a88394467597d3c42bb3e23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/workspace/harbourview-office-relocation/plan.md", "bytes": 3066, "sha256": "ede460c0ae0b1089fcc3db9aafd409e407ca51f947a4454b76a81ecef7a20087", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/workspace/harbourview-office-relocation/plan.md", "bytes": 3066, "sha256": "ede460c0ae0b1089fcc3db9aafd409e407ca51f947a4454b76a81ecef7a20087", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/result.json", "bytes": 903, "sha256": "c58486ca0cb308e2001919e9d073733da070c5dad4966b65a2d4b02bfa5b2d8e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/score.json", "bytes": 2903, "sha256": "132f4c9a0ae47b4b2b982dd3f4cbcce6b0dc630a3dc12e9361268b2e9441bb24", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/metrics.json", "bytes": 2366, "sha256": "85357f36812fddd2b77a15912eb374dbc75a15a76344a19ea8bc99df9f795f67", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/facts.json", "bytes": 11254, "sha256": "20fb0ed6af3a2c424baed25b47e6518a14fa43ee595ade4042ab1b818416e997", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/codex-inline-judge.json", "bytes": 4078, "sha256": "2e4f4a494b36b8cd90545267c0f796a3ef7ca6d10319039f4e546352689a0e31", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-gemma4-e4b-q4-2026-10-22T04-50-27-157Z-x000/postmortem.md", "bytes": 9573, "sha256": "c3628b11b3c4b2d6ca03c28f7f279312fae6ebd8eca331fbb8267964bb8b04bd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/771632c70d.json", "bytes": 3658, "sha256": "095fe023b47b7310e64f2283cc09588af34c0115d6004e7a6a415b7c47d73175", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/workspace/harbourview-office-relocation/plan.md", "bytes": 3368, "sha256": "44f6af0a2872587b6a075622421b89c875373abb33533ed50c1b36dad25fb81e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/workspace/harbourview-office-relocation/plan.md", "bytes": 3368, "sha256": "44f6af0a2872587b6a075622421b89c875373abb33533ed50c1b36dad25fb81e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/result.json", "bytes": 903, "sha256": "d8c3fe3754af6ec99b5f2745dd045722598d482f8c500d6927e240f1b806528c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/score.json", "bytes": 2891, "sha256": "11b20318c08e62bb7e65d784813e21da86223e069a0b0204069b1d75a3b53ddd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/metrics.json", "bytes": 1303, "sha256": "c3ef79a0eacc4c38907a279bf85962840fa9b2cf6233aacb611d575c419d305c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/facts.json", "bytes": 10081, "sha256": "6e110f5fd63ab04f3e6eace228f258eba6a5074b5a117303e0b28323464705b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/codex-inline-judge.json", "bytes": 4110, "sha256": "563c167e036c872df2763a0aedf3a8bf35eb35b8c2b6b4ff9f573473bb9bdb18", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-qwen3-5-9b-q4-2026-10-22T04-51-58-530Z-emyx/postmortem.md", "bytes": 9797, "sha256": "2e8e02cf27aed4eb06f3272dcc10006ebfbcda0aef165e09eacef9cb43fa7f3f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/05576a6a76.json", "bytes": 3658, "sha256": "0572b9ba2d04e6461b95bf6212b98dae3408511857a979a894ffc9619499a0da", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/workspace/harbourview-office-relocation/plan.md", "bytes": 2668, "sha256": "c329fdf44a00b5273e7076e8c8550c6cefb015ce53c07667ddc6a4e6ec76d578", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/workspace/harbourview-office-relocation/plan.md", "bytes": 2668, "sha256": "c329fdf44a00b5273e7076e8c8550c6cefb015ce53c07667ddc6a4e6ec76d578", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/result.json", "bytes": 913, "sha256": "9902afb6efde4e7fc4321fffb172a09b3e1db2188cb53b8e65e8d25c273ac427", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/score.json", "bytes": 2906, "sha256": "8f4e9c0238447cdb50ed2ff3088218a898a2ac140d85f69d540fb2f04a8189c3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/metrics.json", "bytes": 1306, "sha256": "2d2efff32d1df2aecf42cc0ea1bc3369f7e77e0663965149a0cb826c11c0fadf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/facts.json", "bytes": 10139, "sha256": "5ded97d6eaeb00e5471c8e202c16594366d64dbfb8f361959e4567cc5f7770cc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/codex-inline-judge.json", "bytes": 4146, "sha256": "e5124d29c62953c012e5f6670de259ef843b1f804c9fa6fd177e0b6fb9de9f64", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-ornith1-5-9b-q4-2026-10-22T04-53-44-267Z-35fw/postmortem.md", "bytes": 9758, "sha256": "9f3f8a3f28a2bfc28eaf0011fdd37efeeb133789ec0585b5b76bc7f3f647ba05", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/b3bfdf6244.json", "bytes": 3692, "sha256": "58b1d4904947b695e14f6efbbd4afa248b90f9ef419e2caa4a6fd6ab356305cd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2713, "sha256": "007f2756447b5bc7ceb23097f03657ac96bf51a2e8fe2d0f46f2cc1388cb15e5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2713, "sha256": "007f2756447b5bc7ceb23097f03657ac96bf51a2e8fe2d0f46f2cc1388cb15e5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/result.json", "bytes": 1246, "sha256": "58edad7884da934cc12a5357ff45a78af177747168bd1432bbdcbd4f0a1b82b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/score.json", "bytes": 3308, "sha256": "3e5da28e64296d73b5e451651c4d1ee83b232172973cba90d08172ec9c79738c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/metrics.json", "bytes": 2402, "sha256": "d395164e5f29f68a4c6216b61e02a1ab3d80d0f313afc03c2edbbeb0fa431589", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/facts.json", "bytes": 13878, "sha256": "5e36603affa74850a4cbd291f8bb1c18f8e0ab9bc3776a5846524b9d099663d3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/codex-inline-judge.json", "bytes": 7071, "sha256": "a4cb0bef634c6b49b86ff46d88c5588ab625bfaaecfa6776d50d53af30c9696e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-underdog-saluki-27b-iq2-mix-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-underdog-saluki-27b-iq2-mix-2026-10-22T04-55-40-509Z-r7mf/postmortem.md", "bytes": 9511, "sha256": "66ea7cdd28a22a206a5c492219ccf94b68d49c098c75bf3c0bd428991aea58db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/2a73de31f0.json", "bytes": 6416, "sha256": "b67b8f77728bc49a69258a0b7d6b20b5ee72b9f8df4c92ab7827b18fe01cfdab", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3213, "sha256": "0f5d4927c5d7c5fd1a491fd1cf87e2fcf147d426cd77422ccfa4ce994c6e6510", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 3213, "sha256": "0f5d4927c5d7c5fd1a491fd1cf87e2fcf147d426cd77422ccfa4ce994c6e6510", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/result.json", "bytes": 1182, "sha256": "eaafebad60b6f4020b7ebb10312676f7b24c098dea770b37f905424139312833", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/score.json", "bytes": 3306, "sha256": "3d6b73395571e83c674185f1dd85805152992756f8fef428eda874fe5e967d15", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/metrics.json", "bytes": 2376, "sha256": "132506ff5097792290225730cae9fb02bd30ef5638788c184806b6bf4605bdf2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/facts.json", "bytes": 17179, "sha256": "eea87cb47b551b858c5da646d8b437da4e0148415ec331b2bf4c8bdf7c31c1d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/codex-inline-judge.json", "bytes": 6694, "sha256": "972a51b7900bbb77f1d17cf4af568a5d56fde4a074af4d451e1eef5ef9000e78", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.8-27b-q2-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-8-27b-q2-2026-10-22T05-01-16-007Z-nuv5/postmortem.md", "bytes": 9715, "sha256": "ee05627936ca0cdbe3571e22e0d4a014cc72d941307cfdb78d9c30fd0fbbd1b5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/f8c042a277.json", "bytes": 5987, "sha256": "b6fd3d6f018f9d23e972b999b3a4c89c7afe09607530097e52d30c8835d2ffc9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1797, "sha256": "f2b20f2090b5621061234c3c35da86a966df9d93bd8a572dd356a7fa0c6e5d65", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1797, "sha256": "f2b20f2090b5621061234c3c35da86a966df9d93bd8a572dd356a7fa0c6e5d65", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/result.json", "bytes": 1176, "sha256": "ac7eb21e93b1a708e229e585462102900e8a2a73d4522ed6f8213be887943ea5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/score.json", "bytes": 3287, "sha256": "a3ab00daf6065877ab36642eff45ab06084500661d7c6007fc76af8e9b662f5a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/metrics.json", "bytes": 2374, "sha256": "de777c357de41f1208257c4e7bfdd04094f5ecc014567609ca3409e259a489de", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/facts.json", "bytes": 13688, "sha256": "d780d2555be99d26c0279d6674f580d4da8b0f34e267f7537ea56a29889d82c5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/codex-inline-judge.json", "bytes": 6666, "sha256": "57ee823f0f39858ad18f5531c6f25dfd2fae1dd6fb072dcafc7021f336163625", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-12b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-12b-q4-2026-10-22T05-25-56-780Z-g8e4/postmortem.md", "bytes": 9619, "sha256": "01aab2012bc861f7692e47e5eb1d454b5b83d34cafe75ecf7420e922ae164076", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/7711e0abf1.json", "bytes": 5954, "sha256": "bb83682173755d96b282965f49f19fd18fdd368dbd228f7136cdcf248c5c4030", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1357, "sha256": "c6a2da3da7ddad59d81f9ff02890e98e909603ca3efba768744d4212c8fdef20", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 1357, "sha256": "c6a2da3da7ddad59d81f9ff02890e98e909603ca3efba768744d4212c8fdef20", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/result.json", "bytes": 1174, "sha256": "254cd6ffefe5d73c87b244c85c1fe204641a0fac436f6a0b0d2142a28c68aa8d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/score.json", "bytes": 3285, "sha256": "576995fe1bfe520adc1c1dc766eecb544b6dbd7ae4e6f0c03764b38b81ce8299", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/metrics.json", "bytes": 2372, "sha256": "09e32fa13e2715a6afdd7b918974f60bf9f589df7c0b762e74724f082295abcc", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/facts.json", "bytes": 15918, "sha256": "512c42d307e71a54b13ae9545acfc2e7897ead909d0264042ad0a4f7bb81ad3e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/codex-inline-judge.json", "bytes": 6683, "sha256": "d36aa0cf61db861486f4fbfb81c999fdc958d0b845c8e4032026042df1b5bf4e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-gemma4-e4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-gemma4-e4b-q4-2026-10-22T05-28-46-206Z-zftm/postmortem.md", "bytes": 9569, "sha256": "4766af7dc596b487d72043103d503b68fb975f23bda47be624b407e76048bef2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/9e719b1005.json", "bytes": 5954, "sha256": "bda8d5e94453d7e65f0689fd7570df9868738f2c0f50520e1f60882306672539", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2451, "sha256": "5471efff04bbfc782c091bde42fddcdc91013d16b1851c95c59ccc129ff487f5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2451, "sha256": "5471efff04bbfc782c091bde42fddcdc91013d16b1851c95c59ccc129ff487f5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/result.json", "bytes": 1627, "sha256": "7299d28aabda5105ba001a04dddf7662edaf3cffbe10a932450722178aebb6ac", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/score.json", "bytes": 5519, "sha256": "53758c098b738d32ef474e3a66b8080b09d63f656d120dbb24e388a7d833dad6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/metrics.json", "bytes": 2373, "sha256": "63a2522c91b7020ad3d46cb2c01f3fdb41e57a19ac9144a285704313136ec391", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/facts.json", "bytes": 17423, "sha256": "cd3134d985032ec6942de802d957f8d32d61dee98322991bb438b5a83e645b29", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/codex-inline-judge.json", "bytes": 6650, "sha256": "83fe5f256f1aa5dca6fcf42b24f096f297e434d87d007cf4110ccd553d563d52", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-9b-q4-2026-10-22T05-31-20-017Z-dt2c/postmortem.md", "bytes": 11593, "sha256": "45705d3a4d5df55bf05242641011e9bb96b0d2e5245ad989360fe749c055a7d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/4bc61b4479.json", "bytes": 5954, "sha256": "f739aca08e4de474086103b8d06366bd1b0bca11c7207e982e1d249946a8a4c3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2705, "sha256": "f9ed871272ae7419fc4388f4d0ed73c14035a426f9b8ad2384e09ecb0eb273a5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2705, "sha256": "f9ed871272ae7419fc4388f4d0ed73c14035a426f9b8ad2384e09ecb0eb273a5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/result.json", "bytes": 1896, "sha256": "f57b2ce1e720046f3b785783720f18e81e64c29c03e27711955d74e94796b7ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/score.json", "bytes": 5752, "sha256": "928d4b13620454402936a22429c4ea2d3d8069338bc9bdd09c530423766f6e6a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/metrics.json", "bytes": 2384, "sha256": "3531aec13507a221af0c5dc083b06cfb2235e77e382cc913dfd3067191ebf6b8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/facts.json", "bytes": 19448, "sha256": "a090d314e4cb236758f4c89cf8baca76cab026288102394d1b468889d720fd58", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/codex-inline-judge.json", "bytes": 6716, "sha256": "72d60cb33e175d6bdfa648b7e8725c7812ee5f29ddf525d74dbb559566fe1c62", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-ornith1.5-9b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-ornith1-5-9b-q4-2026-10-22T05-34-34-788Z-b8ly/postmortem.md", "bytes": 11510, "sha256": "96237cfc4a8a7cfe15342a0274ba334867c83a3bf72cfda9c30604a953433615", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/615f79172a.json", "bytes": 6020, "sha256": "c5c4a45fb727575ad0dd19f0dec4c9620bf5fe6dcd431be3ac911448124d636d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2181, "sha256": "026fa5b9aa2a6c3ccbdb788ee0edc358ff6d6216bd20a7bd0e06682b6300e837", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 2181, "sha256": "026fa5b9aa2a6c3ccbdb788ee0edc358ff6d6216bd20a7bd0e06682b6300e837", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/result.json", "bytes": 1173, "sha256": "7ca0f6edeeb9714e0b7a902bbee30d2f86272b66fdcfde46c76de91b6ad03f00", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/score.json", "bytes": 3277, "sha256": "d0c5f9bf7997dc6cbaab037c0c5b128706bbe937598106c12378917413a59710", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/metrics.json", "bytes": 2367, "sha256": "3389b6497db5566da7d8471add7f8f2882c829a04783e0c7f157f90412a32560", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/facts.json", "bytes": 13966, "sha256": "fb8a3a3421c3d2686929a90a3286310315c317cf8e02ae802559af72ed0ee698", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/codex-inline-judge.json", "bytes": 6651, "sha256": "2fd109a07adc03ddc5aa0f39a06604c5f3beffbf075994af286875b18ff4972d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/r3-qwen3.5-4b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-qwen3-5-4b-q4-2026-10-22T05-40-00-220Z-bp8o/postmortem.md", "bytes": 9524, "sha256": "7a491f2f375545eca90f73ef4a82c303c18a25b098f69d2559f0c132afcfe026", "verifiedAgainstRecorded": false }, { "path": "evals/runs/small-model-core-continuation-2026-10-20/advisory-document-evidence/5bbfc53303.json", "bytes": 5954, "sha256": "27ea15f1d5714a48eed8af7f90f04bda342ced7563747f60ee981be439459509", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/workspace/tic-tac-toe-game/index.html", "bytes": 3750, "sha256": "dc95ea7ea8e346e11b3cd1029c72fa1615ef18e5aea1e422c6f61740f6a770bb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/result.json", "bytes": 868, "sha256": "456a7c41d4a4a7887b5d31cb0336cfcb48b851620b81acfe5e1ad18badda1e81", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/score.json", "bytes": 3307, "sha256": "54941a5b25d5c86e4d0e5a60747b10f190a48c60a4f0105f2b5036b9d22c552c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/metrics.json", "bytes": 2382, "sha256": "014d458d087ba0c7bcda2305f546bdb40045ac14b2927d7d287cc024d1e5fe6b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/facts.json", "bytes": 12662, "sha256": "c4e2c4884cb35f020b52daf64017030d7c8f55a225fbd8a5c97de809f0ad435b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/codex-inline-judge.json", "bytes": 2910, "sha256": "10746b7dce0ba3f6a9884148ccdc13249a72062bc37eda1d2ae5c1b90575196a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-20T16-55-19-362Z-inby/postmortem.md", "bytes": 7854, "sha256": "3a91a33dec7486f1adf4721a685af9053d8a6f48d0e61eaf93723f100a01b120", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/9cb7a25c13.json", "bytes": 2135, "sha256": "fdafff83f2224b543e871c881eb3dfb07e79d9635966132d6b59f5a7cde72687", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/9cb7a25c13-desktop.png", "bytes": 7721, "sha256": "dfb864849e9c3a24d0d5ee2d45b129e3b540101dff8e1559de83ca4017e16ec7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/9cb7a25c13-mobile.png", "bytes": 7627, "sha256": "4ead9379bcffb603fb7d02f2f3b261e4b79e02c6045778749d96e34f7132cbb7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/workspace/pet-shop-website/index.html", "bytes": 7448, "sha256": "499cfc3923b268490ca59b89799fbebcaccfea821a065bf7590f36a2d8c0c053", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/workspace/pet-shop-website/assets/logo.svg", "bytes": 2804, "sha256": "813a4a423c182d02c542236a5eaf9d9075025bd89f668820cac4ba6019afbf08", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/workspace/pet-shop-website/index.html", "bytes": 7448, "sha256": "499cfc3923b268490ca59b89799fbebcaccfea821a065bf7590f36a2d8c0c053", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/workspace/pet-shop-website/assets/logo.svg", "bytes": 2804, "sha256": "813a4a423c182d02c542236a5eaf9d9075025bd89f668820cac4ba6019afbf08", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/result.json", "bytes": 1408, "sha256": "c1d8e7f2585993b8792ee68a85a0c335d4218e353ce99c602a15880e6a2b574a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/score.json", "bytes": 5847, "sha256": "d3c2e15db4636dbfcb5f28bf203cc09f60299a24ca701540ce0f3c04fad8ef87", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/metrics.json", "bytes": 2396, "sha256": "b2bf6cac140fa32baa8d0d89566822825f6b334e5e12725737b3f3eb58326521", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/facts.json", "bytes": 16302, "sha256": "785b0bae0db109508930d8daff285975e771cf4273f68583956407872d285cc1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/codex-inline-judge.json", "bytes": 3922, "sha256": "547ad6ae47b2c687caddeaaaeeec1f13c5c9c8371903816f9d12233fdf15574d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-20T17-02-05-215Z-qsj4/postmortem.md", "bytes": 10342, "sha256": "10a325205a1a5bd3290def11f11fa972200f905cc09c227ab91eeff17eb0e925", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/3aeae200b9.json", "bytes": 8278, "sha256": "aeb53f1683cc0c8c598288673f6bf3112df5234bbdcb97db1808673ce19cf7a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/3aeae200b9-desktop.png", "bytes": 101626, "sha256": "736e85d34994730756db08a2db83a865c6ea72b0f10a8c71a0ca4645d56cfbe9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/3aeae200b9-pets.png", "bytes": 112949, "sha256": "528fed00e3967e3416e8a7aeaeb3d69348b147ab2d88ddc87b4832e4c42e616a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/3aeae200b9-mobile.png", "bytes": 44303, "sha256": "1d98774f4cbe3c04e1bb96da054df469094c6b2a7d6573deadc55b59923fc3b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/3aeae200b9-mobile-pets.png", "bytes": 60728, "sha256": "a59227c5ae6da4421cd612ea4909a5a73354b163484332504feef5275e8ab4eb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/workspace/tank-combat-arcade/index.html", "bytes": 9425, "sha256": "64973affe3cdd17b02a68a7809301f9663e57da6afb83831cf9a11c96e0e17d9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/workspace/tank-combat-arcade/index.html", "bytes": 9425, "sha256": "64973affe3cdd17b02a68a7809301f9663e57da6afb83831cf9a11c96e0e17d9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/result.json", "bytes": 1370, "sha256": "e29138dc4c514db74843a949096ad4183f4237c30412df433a5fcd6bc4a673a5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/score.json", "bytes": 3825, "sha256": "85c3d0e77f1439d1fa4a5877963a849e15372e02353e2b8dffd69d0121c58003", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/metrics.json", "bytes": 2389, "sha256": "10edd7c48cb31b3d99fad7db28d0600aec971ffade1105810d6e4ba2d9f9bff6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/facts.json", "bytes": 14782, "sha256": "371054efa9af9306fb66e6cf5f228a097ce06ad21f1a3b029c8d95dc9dce993b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/codex-inline-judge.json", "bytes": 3388, "sha256": "cb498412e502a2347bb793cf9a3648f0387661e8aa4df68c7b1cf61df8a85ae5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/r1-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-20T17-28-18-651Z-f3rf/postmortem.md", "bytes": 9076, "sha256": "126b9891d36bdd6deeb724b50408b0d07d186038993f5181b206410afee2c0a2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/13d0445595.json", "bytes": 5455, "sha256": "785f28ce6f1352bca664db0fd15091c0f039eb0be1cda5aa54c538ad2b73ba73", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/13d0445595-desktop.png", "bytes": 23473, "sha256": "0bc436d3e091f1d3fc7ada00293b58d3c6d34abeaa4df9ba6f124c34d3860b7a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/13d0445595-after-input.png", "bytes": 23473, "sha256": "0bc436d3e091f1d3fc7ada00293b58d3c6d34abeaa4df9ba6f124c34d3860b7a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-2026-10-20/advisory-browser-evidence/13d0445595-mobile.png", "bytes": 12933, "sha256": "752df4307ebe753079f1c00d9cf52cf28b41bc0112e7d05c259abbf25db627a3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/types.ts", "bytes": 534, "sha256": "c93abe47eb6f8b6406132c2f79cec372b4cd509f1cdc93f28fe81b762dc59035", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/handlers.ts", "bytes": 681, "sha256": "cc929bf59f0723e999e4125e5f15b953469d43f5d4afe479c2a12dfcfbdff62f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/migrate.ts", "bytes": 783, "sha256": "721481f182968502b8f8ad065bf617551601eba6193a95358d750b814b9ab422", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/default/package.json", "bytes": 128, "sha256": "6aeee50e1df303e1c6f34302b42f08639395284c0650a376550390aa012b9727", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/default/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/types.ts", "bytes": 534, "sha256": "c93abe47eb6f8b6406132c2f79cec372b4cd509f1cdc93f28fe81b762dc59035", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/handlers.ts", "bytes": 681, "sha256": "cc929bf59f0723e999e4125e5f15b953469d43f5d4afe479c2a12dfcfbdff62f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/workspace/user-store/src/migrate.ts", "bytes": 783, "sha256": "721481f182968502b8f8ad065bf617551601eba6193a95358d750b814b9ab422", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/result.json", "bytes": 1570, "sha256": "c9686bdad55b823f743d1d69de145d11b28c8af604c5388a6cae51e9761df8b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/score.json", "bytes": 4352, "sha256": "6fdf10467b8bd1441e4dc9c6c119d0b7aa14f5419d0234b6371158d022fe4dc4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/metrics.json", "bytes": 2380, "sha256": "e43fdbabed8a3e53f9415c84e48846ea88e34a398083df3991170be3c676b4c4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/facts.json", "bytes": 15219, "sha256": "c872952d33cac8617466bd716274ec790f6f03059727a24b0c3eceaca9b20dc0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/codex-inline-judge.json", "bytes": 6009, "sha256": "69195414aaa4e9220863f5a713f095f3fea4f56e9e640c040a07f10a7899e810", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-20T18-15-01-862Z-c19w/postmortem.md", "bytes": 9277, "sha256": "4db95a42c7d3c744cebbf4215cf770c0f45a19a51df44264f2cc303dce0dfbdd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/advisory-code-evidence/d86e968253.json", "bytes": 7765, "sha256": "525d64380ad73591f54bbeded6d0f91533196eeae2bba79318f95a5d1c397ce4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/workspace/machine-kata/src/machine.ts", "bytes": 1631, "sha256": "c1c566a30e25e40e239c496e4a66da6e24e25c9f66f45b7bb96304c11eab3bb9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/workspace/machine-kata/src/machine.ts", "bytes": 1631, "sha256": "c1c566a30e25e40e239c496e4a66da6e24e25c9f66f45b7bb96304c11eab3bb9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/result.json", "bytes": 2094, "sha256": "061a9d117d3595ff1cea1dee2bffe137381217944aaba6f3f2b16c0cb16b66b2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/score.json", "bytes": 4064, "sha256": "a12353387f2d84d42f30f80c01065bff9e4217d6f548b9427270d0393b1570e0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/metrics.json", "bytes": 2375, "sha256": "9a21ebf111565dea229642a6664249b93d5e439860f4b672dbbfc5bbc4f67333", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/facts.json", "bytes": 15452, "sha256": "7a95ae159887db6469e003c097f9eb6a0b5e5147d0f52ed5462975a8ac022996", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/codex-inline-judge.json", "bytes": 3185, "sha256": "02f2dbcdfc8cb016816f1bfaafb69f017786bd405591a1c0259645d94e042ba8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/r1-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-20T18-18-38-904Z-nywy/postmortem.md", "bytes": 8803, "sha256": "6fa8707040fd2a7350edb968123b2f422ce14d0eb6c8a15138bdeca120121aa3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-20/advisory-code-evidence/bbdae3e60a.json", "bytes": 49481, "sha256": "f04548b791936e3884da79edbc6005fd7200a9a1dc1dce1a80ed9a14d4c12807", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/workspace/pagination-service/package.json", "bytes": 120, "sha256": "4c5c2652bca3bf0d3a42c8997e49379234b83b6d089fd46bf4c4951fbec107e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/facts.json", "bytes": 10684, "sha256": "5893934c6949436bebdc5803111f08f8a0574c70c73cbb0e946ff1bdd02af546", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-code-evidence/symptom-debug-brief-source.txt", "bytes": 1699, "sha256": "6acdb2ebc8cebb3d022d9de2c4dd7910aac473bf5fe8291b01b6b724fd0e7a09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/workspace/pagination-service/package.json", "bytes": 120, "sha256": "4c5c2652bca3bf0d3a42c8997e49379234b83b6d089fd46bf4c4951fbec107e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/facts.json", "bytes": 10684, "sha256": "5893934c6949436bebdc5803111f08f8a0574c70c73cbb0e946ff1bdd02af546", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-code-evidence/symptom-debug-brief-source.txt", "bytes": 1699, "sha256": "6acdb2ebc8cebb3d022d9de2c4dd7910aac473bf5fe8291b01b6b724fd0e7a09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/result.json", "bytes": 884, "sha256": "4230760e38b14b5e5b4fee6be9915b4ed840a85e5fa7ac4201f6b1d2856ef1ea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/score.json", "bytes": 2818, "sha256": "157710ad8d41880016546a037742a329cdb61b1c0b3dcd84a0f49ed7b5faf644", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/metrics.json", "bytes": 1307, "sha256": "05f55c6f50f1dde0ab925aa40fd7094ff9e51b200f909dfd855dd52fbb840222", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/facts.json", "bytes": 10684, "sha256": "5893934c6949436bebdc5803111f08f8a0574c70c73cbb0e946ff1bdd02af546", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/codex-inline-judge.json", "bytes": 5632, "sha256": "f8885203f147206a280e67591427ea2c0dae163473ceca6235889314fa8f953e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T05-57-06-415Z-5j7q/postmortem.md", "bytes": 8617, "sha256": "7ce5b968e4408e18d4deda22a75cd2a9098a4df7f9cf2efac3dbecc72c7f31e6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-code-evidence/f2a64617c8.json", "bytes": 11282, "sha256": "3f9280f046f56132c6f6898d53761335d22b94f66bb95b29854db38cccf6a81d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 3843, "sha256": "df47c980be08112e1150bee2d5e7c5df9c042455bfec34a30ff53f4f2665616d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-data-evidence/d24aa7033c-diagnostics.json", "bytes": 8141, "sha256": "403cb824e14dabb8dc2802faf46610a76b64bbea3219b6b3b178b321d82ec979", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/recording/chat-events.jsonl", "bytes": 4412032, "sha256": "c2e5c4daf82756f4c7ea73dae54ac626812d131fc0df135468171caec8e75e36", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/facts.json", "bytes": 10682, "sha256": "49b718b73a2761af7aac0f875702a31513d715dc1ab68f59f1f8cf3bdc4c1ea4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/daemon.log", "bytes": 332521, "sha256": "3f9942043fe4d2b0dba22e08a45b28feb6a6298a8713c0264a4b4039d0b1d036", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/log.txt", "bytes": 15582, "sha256": "321e43b97de651e54d3645efc3eb719a93645aa89dcc05b964febb6112c4ba44", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/result.json", "bytes": 1309, "sha256": "621d78a4d89a120e80865fddeeb2fa35887a408a5b00948fd6b11cfb4313b6c3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/score.json", "bytes": 3267, "sha256": "a51f38d4e39fc302989956bfcb349c124c8a17f6bb9a79aa6c09143147317c03", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/metrics.json", "bytes": 1311, "sha256": "2b8ed5f690b045197c5f215b1bf48a03fd1ad3ed4a04a2ee41cc14a9b80814e7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/facts.json", "bytes": 10682, "sha256": "49b718b73a2761af7aac0f875702a31513d715dc1ab68f59f1f8cf3bdc4c1ea4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/codex-inline-judge.json", "bytes": 5779, "sha256": "ce0e7c7e27c1d9c2968f651e1526fee9b265d508a88d25ec98aae3bb7ba8fc15", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/r1-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T06-07-16-309Z-siwd/postmortem.md", "bytes": 11150, "sha256": "b0601ab7fccf4f3fc8994748eb86cfede57af421a4340147855ce73b448d11c1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation-2026-10-22/advisory-data-evidence/d24aa7033c.json", "bytes": 6443, "sha256": "ef4fd41fca5d1e4ba00cb999f397be45adfdc4060c46913ce4f0096ee07c15a4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/package.json", "bytes": 149, "sha256": "0aabbdefb60c2f087f78eedabdcf50516faad0a9e4a2b70929aeca9e4e8977cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/advisory-document-evidence/1f6ebd7088-diagnostics.json", "bytes": 5754, "sha256": "268dbf92240a649cf598940a3bbc8246dee88d0bea19b7ed2a14e34ee050fd36", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/package.json", "bytes": 149, "sha256": "0aabbdefb60c2f087f78eedabdcf50516faad0a9e4a2b70929aeca9e4e8977cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/workspace/checkout-incident-postmortem/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/recording/chat-events.jsonl", "bytes": 1174606, "sha256": "94d2d26b6e232e00171afe8760bad13f3ff98100b0447792473e273dc548f6da", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/facts.json", "bytes": 10318, "sha256": "eb68a88da413ae51a8ebdbeeb68452eb0606b467c449d956580089fcb5381f16", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/daemon.log", "bytes": 185018, "sha256": "de841a76c37f230feeef9ffdbe71f9133db4232853883f436a9eab688d69b23d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/log.txt", "bytes": 13083, "sha256": "33913470c522a4062eb6c29bdef3a09b69d51a4edb7567e746a8acdea293ea04", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/result.json", "bytes": 1358, "sha256": "b81db5c4f5e4e19d8166cd6f8ee919f5f8a90e0597bb123f754d021430ee06b0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/score.json", "bytes": 3127, "sha256": "b8859e3e02527848ed708c7e1672f952c2240f6d822d51e02042064e90ae2cd8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/metrics.json", "bytes": 1311, "sha256": "6b7b722d3977164a099545bde45bd025c31e4a28523cf857f9881f3b7a0f5cfb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/facts.json", "bytes": 10318, "sha256": "eb68a88da413ae51a8ebdbeeb68452eb0606b467c449d956580089fcb5381f16", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/codex-inline-judge.json", "bytes": 5736, "sha256": "4ae313f4a36067fcf8645e0fcb4d4fb44d9755fe0d242c204f1a37adcd56df6a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/r1-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T08-02-02-672Z-js6y/postmortem.md", "bytes": 9995, "sha256": "7fbc8fc9946ae0d7028786392e5e6aeaaec3480cccac9c7b3ea70e5852a5ed80", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation2-2026-10-22/advisory-document-evidence/1f6ebd7088.json", "bytes": 6735, "sha256": "8f6c71f41d70f6a6f2dda94b4cc6dbdccdcea94aa13081d172ddc4e9c8a4f1fe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/runlog.md", "bytes": 1888, "sha256": "c5ab4b3c655ac651a69355d24b24def5c6f58e41fa4de0fd8a2b4c26503d82be", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1766, "sha256": "63870b6a9eca3fd10b021ecb6b9d0262adf4efed0e086f23e20ae2f1d870e702", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/0ef033ed95-tools.jsonl", "bytes": 4071, "sha256": "ded56c33f8e96ec2478b9db000c8c0e7af1145607fd45ccd129e18b04715b5de", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/0ef033ed95-facts-selection.json", "bytes": 4442, "sha256": "6045b0041d8b2fe8c42297f36605b24926e1375a7dc4b292fc5a8bd69bd5534b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/runlog.md", "bytes": 1888, "sha256": "c5ab4b3c655ac651a69355d24b24def5c6f58e41fa4de0fd8a2b4c26503d82be", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1766, "sha256": "63870b6a9eca3fd10b021ecb6b9d0262adf4efed0e086f23e20ae2f1d870e702", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/0ef033ed95-tools.jsonl", "bytes": 4071, "sha256": "ded56c33f8e96ec2478b9db000c8c0e7af1145607fd45ccd129e18b04715b5de", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/result.json", "bytes": 1017, "sha256": "e4d521fa659434c2972d3b3ed13e57a5cd317df8d83589716a5042a18546a541", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/score.json", "bytes": 2970, "sha256": "843b923c7a6d264cddbf2022996ff510fe70615d5a2f3c71e65bdbe6ea8ca7f4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/metrics.json", "bytes": 2372, "sha256": "92456f86ad325b1b0a95cf2c42f4c6fb411dc777567b4a9d4274abf48f8a45b7", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/facts.json", "bytes": 13871, "sha256": "94e80833ee161e32c3e5175d3dc2d7f64e47fb404c064ac08ef40b848de97d40", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/codex-inline-judge.json", "bytes": 9461, "sha256": "23041deb63fe2792c34ffdd3d13e316e413524cb2b642159707f1233f7073fc6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T09-03-37-117Z-ad8e/postmortem.md", "bytes": 11231, "sha256": "751e20dcb05055dc53d128b3e3744c278a210a44a24fb2dfbefff921c40af54b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/0ef033ed95.json", "bytes": 15022, "sha256": "958f39e82f0d2d2754e13c61165c7377eaecf449384e3fdfe032026fd8608d1e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/workspace/harbourview-office-relocation/plan.md", "bytes": 3927, "sha256": "7ddde7a031016bb136e857d61403ff5162e4799f21e9527cd45b455e3e44c718", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/7e3af2da10-selected-context.json", "bytes": 2780, "sha256": "d2d6ea000e47d7549277472cb1e8f6ea88323c00bea964fddc540fef68f62be8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/workspace/harbourview-office-relocation/plan.md", "bytes": 3927, "sha256": "7ddde7a031016bb136e857d61403ff5162e4799f21e9527cd45b455e3e44c718", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/7e3af2da10-selected-context.json", "bytes": 2780, "sha256": "d2d6ea000e47d7549277472cb1e8f6ea88323c00bea964fddc540fef68f62be8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/facts.json", "bytes": 10350, "sha256": "892bd5c2558db221a078e93e902b509dabd35ae6694b8687b257e2c958624e74", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/result.json", "bytes": 925, "sha256": "6485673770b2f0b1eaddd9905ba6bac6c2a7b3952f2e48c5954b4d8497701bdf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/score.json", "bytes": 2913, "sha256": "46187e8f9835faf86b1b28d775944779eafe1ddd8fd24f4ec9c18fa2180fb1b9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/metrics.json", "bytes": 1307, "sha256": "31fee6db06bf6f98948068b222fc0261fbd596ef4f4d07f11a022a35e93d23e8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/result.json", "bytes": 925, "sha256": "6485673770b2f0b1eaddd9905ba6bac6c2a7b3952f2e48c5954b4d8497701bdf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/score.json", "bytes": 2913, "sha256": "46187e8f9835faf86b1b28d775944779eafe1ddd8fd24f4ec9c18fa2180fb1b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/metrics.json", "bytes": 1307, "sha256": "31fee6db06bf6f98948068b222fc0261fbd596ef4f4d07f11a022a35e93d23e8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/facts.json", "bytes": 10350, "sha256": "892bd5c2558db221a078e93e902b509dabd35ae6694b8687b257e2c958624e74", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/codex-inline-judge.json", "bytes": 4644, "sha256": "9d64a77550bcd5cc1d81bc0a6a7976b3efcc62d3ed7080a95bbc3263cebd7aa3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T09-49-18-715Z-h0ld/postmortem.md", "bytes": 10798, "sha256": "2a833a4ae428f84919ab3fa98d6919840d816053745d956a4a29bd3390bea068", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/7e3af2da10.json", "bytes": 5161, "sha256": "d3ea90c647340466ad26207fbaa1de2643dd05b3331569147d202a1ac82f9a5a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 5161, "sha256": "4b298de675c100842012a30a84b293789f20e5f9734a5a37eb99bfa182942bff", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/ecc910ec66-context.json", "bytes": 8011, "sha256": "3becdb2320db6b663b0cf0c0a6b9eaebe416f5b1ddd365154640b922a913a912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 5161, "sha256": "4b298de675c100842012a30a84b293789f20e5f9734a5a37eb99bfa182942bff", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/result.json", "bytes": 2215, "sha256": "b33ba1f9381a39c04255582f6edaf908b2d3989422123af09d073ca806d664df", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/score.json", "bytes": 6463, "sha256": "ae9fde59c2cdb714afd9b76b6dafe961d17acc824b4827f51871247b0f97402e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/metrics.json", "bytes": 2379, "sha256": "84d37b2f78e4ebbf5679b483276dcb4d230a5be196e798c2616fae95a5aed807", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/facts.json", "bytes": 16286, "sha256": "42a5c52a8de39dcce74c9614c6dfb9e1990fb6907b9519e41b01fbbd51daf6ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/codex-inline-judge.json", "bytes": 7269, "sha256": "22b1dc114fa7e6e943cef671b0463c1c0343b1e5c203af2f5acdc3478867b994", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r1-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T10-02-22-475Z-pbet/postmortem.md", "bytes": 12572, "sha256": "b3253000fa35e988387a9e3c000b735cb7253293c6e659a5e6a6f37840d14bd9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/ecc910ec66.json", "bytes": 6118, "sha256": "6eee98f361f69e694a4a1c7531273be3fb574cd7982dbd0b2bf75d23fe112d0a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/workspace/tic-tac-toe-game/index.html", "bytes": 2262, "sha256": "bca3c4304e654112ca6ee8305cbaa2d681b898580588d3148ce4f29492c479ae", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca-context.json", "bytes": 4226, "sha256": "94a2bfa320d61afe1ddf1a942879f92c43fdfc021b084843df88851b15715a64", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca-replay-selection.json", "bytes": 4849, "sha256": "b1d71409546c37592248e06a8c614b26ed9ae5732ff98182b8aa27af6aae6069", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/result.json", "bytes": 883, "sha256": "f2f4174bab2a80fb18c01dee5eeccff9f70c7714c5123c55138f0cbd6bb402d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/score.json", "bytes": 3315, "sha256": "fb5badeb5b1b24c76bcac0817e070a16cc9bbd10d118f13ac41a33b3bbc49257", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/metrics.json", "bytes": 2380, "sha256": "4db639b3ff6f3ddca5d0476713a194a4ecb82a51f684408565a6e1f683db3137", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/facts.json", "bytes": 14329, "sha256": "27ddbdec20c29865aac53147275a10b7754a94b7dd05dcc6ce9a7a3459ae55b6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/codex-inline-judge.json", "bytes": 4140, "sha256": "a083208c4765adb674b88b450c8e3ab2dccaa0c1d9d76bcc2bc2ef146988ec86", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T11-05-01-736Z-x8l4/postmortem.md", "bytes": 10630, "sha256": "b49e4d5b778a9e1209c709f678aa216512cb1c1aebcfd73cbf164df44436dfa9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca.json", "bytes": 5979, "sha256": "b5d5775f881f8768722e2f8719413dbd31415e46771462554b364b664200477c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca-desktop.png", "bytes": 9425, "sha256": "e1b8e7973ca644ba426ac5b76c026e947147ebcc955e1075814d8c24128097bf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca-winner.png", "bytes": 11720, "sha256": "9c2bbdf277509336d39f082afcf99d547eaaafecfc7f3c125d646084dbede3d4", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/ad447ca5ca-mobile.png", "bytes": 6975, "sha256": "e0be2f6ea9b49979f4790892b9ba2d840bf210266ab23650bdc19a789f946d38", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/workspace/pet-shop-website/index.html", "bytes": 6133, "sha256": "36c30eefb45cf131b5109552c4b73c36b8181590e47ae62099702c9433d173e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9-context.json", "bytes": 3851, "sha256": "19b003e2c047047a30cfb58f13b880b3e49acd5c2342d0dbbdaf77f1f83c6b06", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9-replay.json", "bytes": 3167, "sha256": "565ee7e4a701848a5990e3b321ef43094c5535cb9062b22d6a13260d9ea3dfb9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/workspace/pet-shop-website/index.html", "bytes": 6133, "sha256": "36c30eefb45cf131b5109552c4b73c36b8181590e47ae62099702c9433d173e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9-context.json", "bytes": 3851, "sha256": "19b003e2c047047a30cfb58f13b880b3e49acd5c2342d0dbbdaf77f1f83c6b06", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9-replay.json", "bytes": 3167, "sha256": "565ee7e4a701848a5990e3b321ef43094c5535cb9062b22d6a13260d9ea3dfb9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/facts.json", "bytes": 13240, "sha256": "2715888c02460022cf2d9e036c0035293028664e91462ce9761a0b4650a8ce20", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/result.json", "bytes": 1171, "sha256": "d6ef6bcbfc847fde5d2acfb3ae75a40ea31d9082aadb4f818005f982a4129f76", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/score.json", "bytes": 4376, "sha256": "40cb50fd5607f25deb690fff28d822fdafe18e1f1cd85ca40edab5dc06a9abc2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/metrics.json", "bytes": 2383, "sha256": "49b57fdfe210e5751820aaee3b9904eef681599dfb411a15c6bad23ec720aeec", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/recording/chat-events.jsonl", "bytes": 3823021, "sha256": "9d64eb3e54810fed0910b79a257d99c9f853cbe11463954397b5453d17ce26d8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/result.json", "bytes": 1171, "sha256": "d6ef6bcbfc847fde5d2acfb3ae75a40ea31d9082aadb4f818005f982a4129f76", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/score.json", "bytes": 4376, "sha256": "40cb50fd5607f25deb690fff28d822fdafe18e1f1cd85ca40edab5dc06a9abc2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/metrics.json", "bytes": 2383, "sha256": "49b57fdfe210e5751820aaee3b9904eef681599dfb411a15c6bad23ec720aeec", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/facts.json", "bytes": 13240, "sha256": "2715888c02460022cf2d9e036c0035293028664e91462ce9761a0b4650a8ce20", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/codex-inline-judge.json", "bytes": 4126, "sha256": "bdc61ef1d324b611ed72ef8da80bc04e5e24563a3f81c83d552a8d31b8cac359", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T11-51-26-325Z-3q9x/postmortem.md", "bytes": 13168, "sha256": "da199d7ef0cee54b96ce078467f5d05beb01028dd39cb71a5fa2cc4ff51d2fc6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9.json", "bytes": 7738, "sha256": "7ae6de96913636d2adbeb982fdee92ca0cee1977fbe84caed2e93b26dffb6736", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9-desktop.png", "bytes": 28892, "sha256": "d60ad73daeada52d9aa6520566132a10909e71b7f912add59dda7e7f4f4b5704", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-browser-evidence/564f2b8fe9-mobile.png", "bytes": 33503, "sha256": "6291e1db6f4c1f5e71c046ecdacbbca223c8fcef8f2acf239961ee0ddee014e6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/default/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/default/package.json", "bytes": 128, "sha256": "6aeee50e1df303e1c6f34302b42f08639395284c0650a376550390aa012b9727", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/default/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/shared/About this library.md", "bytes": 1170, "sha256": "5b3303e412dd0049ba460d6a45eff4f0c20277d02a87f3ec38de3a6b1efe8911", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/tank-combat-arcade/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/tank-combat-arcade/package.json", "bytes": 139, "sha256": "2ed04a3ff568417303b67d7dd0f8a171fa533ea024eb9207c9dd1e41c9d3c94a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/tank-combat-arcade/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/783572a4d1-diagnostics.json", "bytes": 7382, "sha256": "77df030a0ee339bdfe1652beb817d7844f827e7470a51be9c254415c6d6ac168", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/783572a4d1-conversion-metadata.json", "bytes": 3379, "sha256": "7e7079ee03b1c50eb37320516e6c67f9d11b110362a1935b2ecc96a738ffb85e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/default/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/default/package.json", "bytes": 128, "sha256": "6aeee50e1df303e1c6f34302b42f08639395284c0650a376550390aa012b9727", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/default/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/shared/About this library.md", "bytes": 1170, "sha256": "5b3303e412dd0049ba460d6a45eff4f0c20277d02a87f3ec38de3a6b1efe8911", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/tank-combat-arcade/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/tank-combat-arcade/package.json", "bytes": 139, "sha256": "2ed04a3ff568417303b67d7dd0f8a171fa533ea024eb9207c9dd1e41c9d3c94a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/workspace/tank-combat-arcade/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/facts.json", "bytes": 12505, "sha256": "adc0b0faeb025c1e86bca597e55f23aab7782445c1f649837720b7f6ceae284c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/recording/chat-events.jsonl", "bytes": 11343073, "sha256": "762ea3a721b749dddca0a4cc0836278e8e895cb407ef4153b04bb3e5e17e3d7a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/daemon.log", "bytes": 417855, "sha256": "a2718ec3b4e899500abbb4519f9c4a1e04d034eaafaefe3fb3d00c1945f81bf6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/log.txt", "bytes": 8139, "sha256": "55d0af843cfc431de0a0da3e4d2afd9cd03a77b3756a1fe3bd140cfe5822081c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/result.json", "bytes": 1225, "sha256": "4c66884ad6cc40d19eb02a656ef7edd33ecc77cd1d5759964e8a3221bb3b7f18", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/score.json", "bytes": 3116, "sha256": "6c02d5c435572437e49cec511c31230ac1915791c780b8158e109ed71238b703", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/metrics.json", "bytes": 2383, "sha256": "05d0af10c4e268170d83fa70601697873a6090bb125f7dcd521bcaea86f67704", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/facts.json", "bytes": 12505, "sha256": "adc0b0faeb025c1e86bca597e55f23aab7782445c1f649837720b7f6ceae284c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/codex-inline-judge.json", "bytes": 5302, "sha256": "aa8156eaf3e9bf2271fd1e2b7ae4f6aa811222c16cd1e8d014082b87f0d8f8d8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/r2-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T12-34-11-752Z-xmfu/postmortem.md", "bytes": 10536, "sha256": "0f97a860dc16d233c2907e374d677230bd172ff5cd9bac39eb7fbb04df0cf862", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation3-2026-10-22/advisory-document-evidence/783572a4d1.json", "bytes": 5767, "sha256": "8872439efca75a97168aca252618b574074a6b5653fb1688b026eaa2748044d9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/types.ts", "bytes": 750, "sha256": "a13fc7a9de53f4b057bd74fe4df777660fe68852110cb0a61c094cbff4f0aaac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/handlers.ts", "bytes": 892, "sha256": "ded5f052c854c49d01d91d080593959ceeff3ad87a26c2f8557207e658640d39", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/migrate.ts", "bytes": 1194, "sha256": "54be43f489b509be9727fa4737a76b0d6396487784e51d4a0eba55d43b3f3ca6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/dd0393ffb3-context.json", "bytes": 9089, "sha256": "a569b1effbb63e6079dfc4d8e0c602a11019b799c3fc8a1395ee0f343b68a36b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/types.ts", "bytes": 750, "sha256": "a13fc7a9de53f4b057bd74fe4df777660fe68852110cb0a61c094cbff4f0aaac", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/store.ts", "bytes": 824, "sha256": "abbcac468914e729812abdb5639e3eccee02ee9e9b949057c92c6f1125c23356", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/handlers.ts", "bytes": 892, "sha256": "ded5f052c854c49d01d91d080593959ceeff3ad87a26c2f8557207e658640d39", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/src/migrate.ts", "bytes": 1194, "sha256": "54be43f489b509be9727fa4737a76b0d6396487784e51d4a0eba55d43b3f3ca6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/dd0393ffb3-context.json", "bytes": 9089, "sha256": "a569b1effbb63e6079dfc4d8e0c602a11019b799c3fc8a1395ee0f343b68a36b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/facts.json", "bytes": 15615, "sha256": "72feda129f0ebd97001c1a0875ef1d09a61eb0b8c7bb4d24979421570b1e6b82", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/recording/chat-events.jsonl", "bytes": 1305070, "sha256": "a45515cbf9cad82878fec760dfa545eb47086e4bc712697c5e485b8dd1d37d0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/daemon.log", "bytes": 201772, "sha256": "2f5999c54190c5c71f92a13868b769a9f380a083df25336a68f103d0c638082f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/log.txt", "bytes": 18016, "sha256": "8f0405ce0edee3e9d8804a7d71565dd94952adab65616abd88d3e6f892dfac36", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/result.json", "bytes": 1679, "sha256": "c00959171b0cd45fb5e2ffe3dce0d93ea0b170223c8348d692e962a75f8028e1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/score.json", "bytes": 4569, "sha256": "95b4ebad1fbb61c03f26ae19aeb01915aee068b322a0739777ec2632be2cce5f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/metrics.json", "bytes": 2376, "sha256": "f0fafa57b0a86faaf58d3c04a4ff6824f5478396fc18b5a08a019e52f14d499f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/facts.json", "bytes": 15615, "sha256": "72feda129f0ebd97001c1a0875ef1d09a61eb0b8c7bb4d24979421570b1e6b82", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/codex-inline-judge.json", "bytes": 7302, "sha256": "eebb6b5456cfb4c05c7e58327ac843b77e8eba2de9f93b42a2c36c297de8d639", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T14-50-20-949Z-8vm6/postmortem.md", "bytes": 10858, "sha256": "eea6efa8d94041f8bb165d457649ebb7955c18a6582fcbb103dc9d217bbcac68", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/dd0393ffb3.json", "bytes": 17379, "sha256": "83034e1b59a3aa75f16703ad4fd4fb6d0bb199b92735008c7f1fa27da23bd38d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/src/machine.ts", "bytes": 1308, "sha256": "ae052e3ea2aebea1f3ab3d9ec7ff4cca38ebe69fd5b99d4aa94c2a29ea09a22b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/2210b3e9b5-context.json", "bytes": 5573, "sha256": "cf4338e83bb16dce08a829a0b53de05efb3a65c02b68263050df3adf5fc9afd9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/src/machine.ts", "bytes": 1308, "sha256": "ae052e3ea2aebea1f3ab3d9ec7ff4cca38ebe69fd5b99d4aa94c2a29ea09a22b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/2210b3e9b5-context.json", "bytes": 5573, "sha256": "cf4338e83bb16dce08a829a0b53de05efb3a65c02b68263050df3adf5fc9afd9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/facts.json", "bytes": 14264, "sha256": "c2ce644ade9c04afa4555a24367a6046b3a68af9f5b12095620eb6c700093967", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/recording/chat-events.jsonl", "bytes": 856377, "sha256": "48c34920e1ebcf35373b39af50c891ac2c4600cae7b92bb9a67b9d034d262770", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/daemon.log", "bytes": 320258, "sha256": "6f06ca10dc1f1459c3064f78f89e669bf3cc17bd35eb294ae73bb700456febf2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/log.txt", "bytes": 12582, "sha256": "b1ce3188cec95239c658cb1f1dfb8f7a771e59b33063b5eaa05897f7482d1fb3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/result.json", "bytes": 2051, "sha256": "b9d5fc8854d03e1f2b647c4233f587409a851dcdc0e74ea60a1743bf99720720", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/score.json", "bytes": 5347, "sha256": "7e031290de7e6293643e9788f79e3787d3783295d5ecdd8e62bb6d75cd33b794", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/metrics.json", "bytes": 2374, "sha256": "285d3970c0dd374c896e36d518fa17533392393c20d30d80725aff577948f8db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/facts.json", "bytes": 14264, "sha256": "c2ce644ade9c04afa4555a24367a6046b3a68af9f5b12095620eb6c700093967", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/codex-inline-judge.json", "bytes": 5142, "sha256": "5fa8da2f68e9305c6238dab6d653645a6eea0f1f1d6eff1624360134cbf97d9e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T14-55-44-658Z-jytx/postmortem.md", "bytes": 10900, "sha256": "60c0665e00e292028d64f008c771acf8dd0ff8310219ac9d851201410335c488", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/2210b3e9b5.json", "bytes": 45009, "sha256": "30de2a8a3c85dcd7e4aaf7ee6e3fcbd9fd6ee32b750e6a3a3a72ce082e1a46cb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/workspace/pagination-service/package.json", "bytes": 120, "sha256": "4c5c2652bca3bf0d3a42c8997e49379234b83b6d089fd46bf4c4951fbec107e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/symptom-debug-brief-source.txt", "bytes": 1699, "sha256": "6acdb2ebc8cebb3d022d9de2c4dd7910aac473bf5fe8291b01b6b724fd0e7a09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/b91a796d68-context.json", "bytes": 1822, "sha256": "b6e7824c96c73853a25ef1dc0684b72fcfbac04fec227b9fab8b1d6f3ef85731", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/workspace/pagination-service/package.json", "bytes": 120, "sha256": "4c5c2652bca3bf0d3a42c8997e49379234b83b6d089fd46bf4c4951fbec107e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/symptom-debug-brief-source.txt", "bytes": 1699, "sha256": "6acdb2ebc8cebb3d022d9de2c4dd7910aac473bf5fe8291b01b6b724fd0e7a09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/b91a796d68-context.json", "bytes": 1822, "sha256": "b6e7824c96c73853a25ef1dc0684b72fcfbac04fec227b9fab8b1d6f3ef85731", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/facts.json", "bytes": 10821, "sha256": "36bf9f9ad2ecc95c7e46593469781b32018ee0c18a7d6db9621bd1f5d959f102", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/recording/chat-events.jsonl", "bytes": 442427, "sha256": "7ad9d52820ac8e9d4239ba87b4f7ceb92a524775231148a870a7acf5bff2b136", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/daemon.log", "bytes": 48395, "sha256": "03a256219cf27a2df9dff18c2232201151589822c8ffcafcb22a32804253fb5b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/log.txt", "bytes": 6617, "sha256": "2734074b825c9b67be979cd1f9be7c64a20fe0fca9d16b5ddf941350bb6d2826", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/result.json", "bytes": 885, "sha256": "c8625c924322913e8990616bb4adea4ab353e410548f5aa6ebd3285b8fe3f7dd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/score.json", "bytes": 2818, "sha256": "ef4404ab024b0b68c96120bf70982392d942dbaa1cdc6d2504b48963ffb7d311", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/metrics.json", "bytes": 1310, "sha256": "ba372a2f5c904d35328decf3f919b10a023fb3d9b6f0d3c819162718fa8e5ce2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/facts.json", "bytes": 10821, "sha256": "36bf9f9ad2ecc95c7e46593469781b32018ee0c18a7d6db9621bd1f5d959f102", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/codex-inline-judge.json", "bytes": 5221, "sha256": "d2b2896091e156ebbf40a296c7953827f5e62dfa7062f4ef00f3710f2a831f72", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T14-58-48-577Z-fvil/postmortem.md", "bytes": 8872, "sha256": "f8cb6643721db42ae73b33434b18060be9066a633e255e1b741ad9f8ca0734d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-code-evidence/b91a796d68.json", "bytes": 12660, "sha256": "19e2af029cf50772912defbb23a0d30a3d7811aa56e64ca117c398e662c92334", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/out/customers.json", "bytes": 1253, "sha256": "852537afaee0188ec58582ccfb5de15d9994aeca68d4bfd83cb801039be0168a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2457, "sha256": "53160f5dc05028655d4f3e127f5bcc25d9d3cea4ed048875ad97e2a641957540", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-data-evidence/88bb410c75-context.json", "bytes": 8484, "sha256": "7ef13a005ebc12e028b607e5a48221361160362a8bf95a6effd71adb8f00ef5e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/out/customers.json", "bytes": 1253, "sha256": "852537afaee0188ec58582ccfb5de15d9994aeca68d4bfd83cb801039be0168a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2457, "sha256": "53160f5dc05028655d4f3e127f5bcc25d9d3cea4ed048875ad97e2a641957540", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-data-evidence/88bb410c75-context.json", "bytes": 8484, "sha256": "7ef13a005ebc12e028b607e5a48221361160362a8bf95a6effd71adb8f00ef5e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/facts.json", "bytes": 12653, "sha256": "13f670c83c86471b954544332f7741a37ae6cd101098abbb76b3c7c8ca43fd9a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/recording/chat-events.jsonl", "bytes": 3985962, "sha256": "60f53aa926007dcda07898104ed0dff751f8a9c8d68161ea10613fef6bfa9f4e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/daemon.log", "bytes": 578008, "sha256": "e1efeae82324a13d6d7a5afd9ff262b389ababa913869236e9ec8ee267ffe7aa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/log.txt", "bytes": 15577, "sha256": "3ad32c3067dc82fdb78249c7e7f6f7d7e028c189aa2d51ac26ba7d883a738c70", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/result.json", "bytes": 1460, "sha256": "f5c99fbd20f18aa10a086b6bd7125bf3892b0d1c94f99c71f8eaeca41ee32ad5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/score.json", "bytes": 3537, "sha256": "39073ea05356933b93182734b09169fddfb7616a88057de74b2f4d1221f34dda", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/metrics.json", "bytes": 2386, "sha256": "68e7f46c5e86c332a202890f25423ea74460345b1d695d779338ddd74d102c88", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/facts.json", "bytes": 12653, "sha256": "13f670c83c86471b954544332f7741a37ae6cd101098abbb76b3c7c8ca43fd9a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/codex-inline-judge.json", "bytes": 6499, "sha256": "9103224f336e5418d0c494530f3934fd9c83fbe91d2ab488e5dd747e4d5f29ea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/r2-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T15-03-46-374Z-ctsj/postmortem.md", "bytes": 10301, "sha256": "c08332fbdf54c8f688d141678e485ae48bae83687494a7a140bbfea21c49ab2f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation4-2026-10-22/advisory-data-evidence/88bb410c75.json", "bytes": 10336, "sha256": "5c09c5ce13d03516a2da4a7c31f47b84d6c01e9690d0df42494576bc0c29436d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/package.json", "bytes": 149, "sha256": "0aabbdefb60c2f087f78eedabdcf50516faad0a9e4a2b70929aeca9e4e8977cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/advisory-document-evidence/37b7545b95-diagnostics.json", "bytes": 10763, "sha256": "cc37fafcf6c70687d8e6518601bde6bd103a62a62b9035abb79b596177024505", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/package.json", "bytes": 149, "sha256": "0aabbdefb60c2f087f78eedabdcf50516faad0a9e4a2b70929aeca9e4e8977cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/workspace/checkout-incident-postmortem/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/recording/chat-events.jsonl", "bytes": 2059224, "sha256": "33114947f1f16bc7d6d0f58bc002abcde04df0f49bb4a11fd5fe5e0a95b1e58d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/facts.json", "bytes": 11609, "sha256": "82ace16203e51342cfb9ec48949b79145a51c568dece5327341f10e8dcc7af35", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/daemon.log", "bytes": 300077, "sha256": "6f12fb1aa4084027aa55a257504812c45e314f41034d71e760c87c82e88519dc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/log.txt", "bytes": 12294, "sha256": "d8f79f59f4cb232dcd5388bd9c83b7f586ff74b6d8de7fb1da0f3bd7d23d2653", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/result.json", "bytes": 1357, "sha256": "38dfa33a15104c4d05079b6e4046bc2e67d2a006933f230945e3a49a8d07e335", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/score.json", "bytes": 3082, "sha256": "73a875cd3388bd2ebef8c59f0b0a1d8fbb4fb778a9a23b5c64c2a25b8ba6e82d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/metrics.json", "bytes": 2385, "sha256": "bf2169fdd06f5b5d689ecda49b6a893764f891964662a02a6fef283eb54b164f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/facts.json", "bytes": 11609, "sha256": "82ace16203e51342cfb9ec48949b79145a51c568dece5327341f10e8dcc7af35", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/codex-inline-judge.json", "bytes": 5909, "sha256": "1c780c68f484e4ca66976ab522b36e843a460585db081f959e7331d34f61138a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/r2-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T15-54-50-578Z-5wva/postmortem.md", "bytes": 10627, "sha256": "009911352b1880a4e6fd9b07a3095696bb07d419253a65fddb80dd8430f75c5e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation5-2026-10-22/advisory-document-evidence/37b7545b95.json", "bytes": 6737, "sha256": "fb4ecf6a28a942f499393eafddd558ddebc300468172ce1ec28112899b79f0e2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/runlog.md", "bytes": 1398, "sha256": "5bb6712e213fd1a5962a47c961e6b88eabfd128d85dd96ead402a1ce4023cbc0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 871, "sha256": "ffc2aeb943252142404398631aae2ddf6fc149dddb098bb199b8f9958b757c51", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/4e011c084f-context.json", "bytes": 4540, "sha256": "16400749f4ebfb32ba7c420259d2b6b7b7123bf1910893419060aed87058fbd9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/runlog.md", "bytes": 1398, "sha256": "5bb6712e213fd1a5962a47c961e6b88eabfd128d85dd96ead402a1ce4023cbc0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 871, "sha256": "ffc2aeb943252142404398631aae2ddf6fc149dddb098bb199b8f9958b757c51", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/4e011c084f-context.json", "bytes": 4540, "sha256": "16400749f4ebfb32ba7c420259d2b6b7b7123bf1910893419060aed87058fbd9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/facts.json", "bytes": 14066, "sha256": "69c2b5279c9b009aeb1acafacd5d55aef7b9e8f2524a358bb631c42242811a6c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/result.json", "bytes": 1016, "sha256": "471aed5cb3f5943c824cf7523f21cd7b7bd823e7f58d93d1d9302074ba6b5ea9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/score.json", "bytes": 2991, "sha256": "f2e00850351676b2dc5dae76fa95d5790cc88a644c3d2f1a54016e3fb30ca5c1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/metrics.json", "bytes": 2378, "sha256": "d421e0f54e6804cd55cc7e20676d49f7dcbc3154161ac5c398924b817ab9048f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/recording/chat-events.jsonl", "bytes": 561322, "sha256": "f1ce1cfd5e7a38b4af5ee0f84f4e2f1482c9d8c22a9ef64d16476c55d905bb72", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/daemon.log", "bytes": 63235, "sha256": "2137143b61806c5bd46b8966158a8c3d8010fb045c3147ba5f8f010942e06033", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/result.json", "bytes": 1016, "sha256": "471aed5cb3f5943c824cf7523f21cd7b7bd823e7f58d93d1d9302074ba6b5ea9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/score.json", "bytes": 2991, "sha256": "f2e00850351676b2dc5dae76fa95d5790cc88a644c3d2f1a54016e3fb30ca5c1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/metrics.json", "bytes": 2378, "sha256": "d421e0f54e6804cd55cc7e20676d49f7dcbc3154161ac5c398924b817ab9048f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/facts.json", "bytes": 14066, "sha256": "69c2b5279c9b009aeb1acafacd5d55aef7b9e8f2524a358bb631c42242811a6c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/codex-inline-judge.json", "bytes": 8929, "sha256": "5c5ae50f4e9317c8aff64fdf9b9e150f27528a402f116eca268709494b3f9dea", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T16-08-03-135Z-ct0r/postmortem.md", "bytes": 10307, "sha256": "c2fdceeb782edf0df05080a76683ba8637d0bb5ba0f0b7728f7bfb99f1fd8e77", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/4e011c084f.json", "bytes": 16083, "sha256": "9a86d9e314d432ecd82dd2f489609c9198400532277e780faf04e051d19fd381", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/workspace/harbourview-office-relocation/plan.md", "bytes": 3479, "sha256": "50d0074928f4e8a07dcceb8911dda083afeb68ff3cdffc2531d53df7ad48ed65", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/550e022825-context.json", "bytes": 3396, "sha256": "e3aa7f56cbb2b0d6e6ccedaef03240cf4a9f5540613484af5a566e6024b2077a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/workspace/harbourview-office-relocation/plan.md", "bytes": 3479, "sha256": "50d0074928f4e8a07dcceb8911dda083afeb68ff3cdffc2531d53df7ad48ed65", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/550e022825-context.json", "bytes": 3396, "sha256": "e3aa7f56cbb2b0d6e6ccedaef03240cf4a9f5540613484af5a566e6024b2077a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/facts.json", "bytes": 10227, "sha256": "c07738f1d8b1a99e2d8b0873192fdfe68492be9b143d66da22677c721b7f4c62", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/result.json", "bytes": 924, "sha256": "65ce762ddc57098e1f58b1fc65200397117de661ddc7006f4c70be58220bc3db", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/score.json", "bytes": 2910, "sha256": "a31fb1fe23be98260cce60ea00cef231b0e1d413bcb90db80211fcebf3da6be8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/metrics.json", "bytes": 1306, "sha256": "ccc9b4946c1552e826bb58cbbb3bbce53dff03b328665efb53151ee48ba79450", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/recording/chat-events.jsonl", "bytes": 476603, "sha256": "46c4d3f7b4e88aaed624dc9d91e756a96611e42586f1256eac15213c5302a69d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/daemon.log", "bytes": 37739, "sha256": "3271b1cdba2b9a53a6c7a6cf9ecae6dce4b7c5c6fc48d97aba90111b99993eb1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/result.json", "bytes": 924, "sha256": "65ce762ddc57098e1f58b1fc65200397117de661ddc7006f4c70be58220bc3db", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/score.json", "bytes": 2910, "sha256": "a31fb1fe23be98260cce60ea00cef231b0e1d413bcb90db80211fcebf3da6be8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/metrics.json", "bytes": 1306, "sha256": "ccc9b4946c1552e826bb58cbbb3bbce53dff03b328665efb53151ee48ba79450", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/facts.json", "bytes": 10227, "sha256": "c07738f1d8b1a99e2d8b0873192fdfe68492be9b143d66da22677c721b7f4c62", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/codex-inline-judge.json", "bytes": 4781, "sha256": "bc3dbb725cec397a602e6afc84c53003b3cb55d6ffdef328e47c81b0b13af6ae", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T16-11-13-834Z-ibch/postmortem.md", "bytes": 10392, "sha256": "45c85ab9435b8395c8a1d7b978f90fa134d828a0cbc1002c02fd7c2bce9db439", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/550e022825.json", "bytes": 8432, "sha256": "ca9390eaa891c8f6a2386e6235e3e49696e441ca99487b102d948354126db23d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 4227, "sha256": "b2a3d8a53b9e31a75eb5844ef0b0a605d31c824cbd4581b3bc98cbd6c6802e97", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/9ce34d9ec5-context.json", "bytes": 8001, "sha256": "ee12394c6218aecf71319bf43dbd91c87c5c112207a30b752e57e7b601aea8d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 4227, "sha256": "b2a3d8a53b9e31a75eb5844ef0b0a605d31c824cbd4581b3bc98cbd6c6802e97", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/9ce34d9ec5-context.json", "bytes": 8001, "sha256": "ee12394c6218aecf71319bf43dbd91c87c5c112207a30b752e57e7b601aea8d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/facts.json", "bytes": 21277, "sha256": "c98cb3714aad9f528d332d039c5d8c55ad0189c86b188e66585e2c1d3d7ba17a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/result.json", "bytes": 1728, "sha256": "59dda715d19670f3459712eff3ee462f0f2669d499f472f9edc3241674ec3c58", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/score.json", "bytes": 4124, "sha256": "814310ff56d625ffcf91a8e0ed95fff91444772c29c60925c268fa3997ef9820", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/metrics.json", "bytes": 2395, "sha256": "4ad8126ed95986ff79c342d7576f3ee7c0c26e1b181e82fae82ee427400198b2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/recording/chat-events.jsonl", "bytes": 10360680, "sha256": "6b49a6d569d484c4ee4a9dc89d870dfd8d02ed673aad9fb8a5997e439198ab81", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/daemon.log", "bytes": 437424, "sha256": "10c0ddb52644928eb9d9e0284a935e0fc74e56fa95013d782f220609c5af2373", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/result.json", "bytes": 1728, "sha256": "59dda715d19670f3459712eff3ee462f0f2669d499f472f9edc3241674ec3c58", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/score.json", "bytes": 4124, "sha256": "814310ff56d625ffcf91a8e0ed95fff91444772c29c60925c268fa3997ef9820", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/metrics.json", "bytes": 2395, "sha256": "4ad8126ed95986ff79c342d7576f3ee7c0c26e1b181e82fae82ee427400198b2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/facts.json", "bytes": 21277, "sha256": "c98cb3714aad9f528d332d039c5d8c55ad0189c86b188e66585e2c1d3d7ba17a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/codex-inline-judge.json", "bytes": 7698, "sha256": "a234283feecf0dcc4ca1c894a2191c424fb59b6cc0b75837dd60c406242b64a8", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/r2-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T16-13-34-842Z-rkll/postmortem.md", "bytes": 11634, "sha256": "53dc59615190b076cef93af3559cfa2449a0a711f9f6d4e7bf19334882d58ebd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation6-2026-10-22/advisory-document-evidence/9ce34d9ec5.json", "bytes": 13093, "sha256": "ec26a40ba228c0b6b60f5767eb13e1c921d8fc8efe730584c77f94a0573c971f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/workspace/tic-tac-toe-game/index.html", "bytes": 4100, "sha256": "150f001b9fc32b9021f15dd10924e08869c39c4d97410552e6ea7dbc32f7523d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7-context.json", "bytes": 3546, "sha256": "703b68b94ffaf3fb35060fc8cec41b3d68852ea587ab5093de7f0ee7bfdd2601", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7-replay.json", "bytes": 1493, "sha256": "9fb12c824cc9305e348dc285c40c0121c228effa2f970837455d0f4f6199afb5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/workspace/tic-tac-toe-game/index.html", "bytes": 4100, "sha256": "150f001b9fc32b9021f15dd10924e08869c39c4d97410552e6ea7dbc32f7523d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7-context.json", "bytes": 3546, "sha256": "703b68b94ffaf3fb35060fc8cec41b3d68852ea587ab5093de7f0ee7bfdd2601", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7-replay.json", "bytes": 1493, "sha256": "9fb12c824cc9305e348dc285c40c0121c228effa2f970837455d0f4f6199afb5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/facts.json", "bytes": 14371, "sha256": "141ab683e26c435f1d44eeb6f69461436d068780884a164db86abc7048965b20", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/result.json", "bytes": 1282, "sha256": "cdb16b861d48d4ccafa9057022d1558e8ab90e20030e0e000395a0d716a1fb80", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/score.json", "bytes": 4557, "sha256": "a1eae7d5fe59651290276392819f7525cc4b5b6751da25da49b7ccded2bd05ca", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/metrics.json", "bytes": 2393, "sha256": "2b8c07b83c69a86fd5c9dce0820010b67dd3b4c86375fcbe9dc6b5121ca4f557", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/recording/chat-events.jsonl", "bytes": 6935573, "sha256": "442c50f943e1f1b693133ebfb449894b276a6e08cbcc04e0fe302242c90f3c33", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/daemon.log", "bytes": 331050, "sha256": "9fd25df76ded1126d0795cbf6f6e3ea94b512e970a8dd386c934cdb3ab742b40", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/result.json", "bytes": 1282, "sha256": "cdb16b861d48d4ccafa9057022d1558e8ab90e20030e0e000395a0d716a1fb80", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/score.json", "bytes": 4557, "sha256": "a1eae7d5fe59651290276392819f7525cc4b5b6751da25da49b7ccded2bd05ca", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/metrics.json", "bytes": 2393, "sha256": "2b8c07b83c69a86fd5c9dce0820010b67dd3b4c86375fcbe9dc6b5121ca4f557", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/facts.json", "bytes": 14371, "sha256": "141ab683e26c435f1d44eeb6f69461436d068780884a164db86abc7048965b20", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/codex-inline-judge.json", "bytes": 4159, "sha256": "47b48d710002ccb963ef8ba6a87ec5ee2bf51fbd4a040c2301b407db03320a6d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-tictactoe/tictactoe/tictactoe-nanbeige4-2-3b-q4-2026-10-22T17-08-05-936Z-925n/postmortem.md", "bytes": 12973, "sha256": "607a299a2baa0c2a7ce20dee300647c00e40ca32e9b5576a1d27b6605a89633d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7.json", "bytes": 8512, "sha256": "0064fe724ab7b0e0e2380499e3800f6373425ff6371caf663ca41312bd781125", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7-desktop.png", "bytes": 317966, "sha256": "a35c070f8b2f269f49505423bcfde75a45423a5720898e6e24d44764c7431692", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/f9779697d7-mobile.png", "bytes": 83853, "sha256": "573a20f476e159cd33266695f3ae9d7a16248026d67738409cce8aee3259d22e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/workspace/pet-shop-website/index.html", "bytes": 9201, "sha256": "b860db2aaa433f5444ba481f631e223cff250cdc2d3ea1d412a39938054eaa40", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/workspace/pet-shop-website/assets/logo.svg", "bytes": 1857, "sha256": "7076051a4c0458bb4f02613de530e4177e8106362c6baaefc90b0bbd7ae2b04b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021-context.json", "bytes": 4598, "sha256": "ee14f6503e4e5b7d0a89190b3abaa8b3bb9bbd07d5770f09b79aa132c25e3ac5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021-replay.json", "bytes": 3462, "sha256": "d8ccd9fbec4227bfbcd40e5d90b5a9e642d9c5eb6faaa84ea451109d3d478588", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/workspace/pet-shop-website/index.html", "bytes": 9201, "sha256": "b860db2aaa433f5444ba481f631e223cff250cdc2d3ea1d412a39938054eaa40", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/workspace/pet-shop-website/assets/logo.svg", "bytes": 1857, "sha256": "7076051a4c0458bb4f02613de530e4177e8106362c6baaefc90b0bbd7ae2b04b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021-context.json", "bytes": 4598, "sha256": "ee14f6503e4e5b7d0a89190b3abaa8b3bb9bbd07d5770f09b79aa132c25e3ac5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021-replay.json", "bytes": 3462, "sha256": "d8ccd9fbec4227bfbcd40e5d90b5a9e642d9c5eb6faaa84ea451109d3d478588", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/facts.json", "bytes": 15450, "sha256": "4a0a1f3a28c853c2d80092cabb1e1e09be0120a9dbd32d4cac72d37a12174f66", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/result.json", "bytes": 1319, "sha256": "a66f3e9d82fde4f5f1751ed6f59e87c39d9404e0672ed9d6582740327386792b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/score.json", "bytes": 4408, "sha256": "d8b9e396433ebecd7ec087e5973ae9e080be2ebd1afa27a419f46b5701787f69", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/metrics.json", "bytes": 2395, "sha256": "e7bddb55b944eb29eafda8ab0c49f9ca14cde4916f95881730664a1f9666e898", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/recording/chat-events.jsonl", "bytes": 4291219, "sha256": "306f4b110e596d397117aef53f5b0595d0a5dfde0a0bcdb1f472bc8d98bb1739", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/daemon.log", "bytes": 389295, "sha256": "881c7e4978586ab2fed3019c89fb9af6be50ad6267ee2e773151c17510aaf2ee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/result.json", "bytes": 1319, "sha256": "a66f3e9d82fde4f5f1751ed6f59e87c39d9404e0672ed9d6582740327386792b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/score.json", "bytes": 4408, "sha256": "d8b9e396433ebecd7ec087e5973ae9e080be2ebd1afa27a419f46b5701787f69", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/metrics.json", "bytes": 2395, "sha256": "e7bddb55b944eb29eafda8ab0c49f9ca14cde4916f95881730664a1f9666e898", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/facts.json", "bytes": 15450, "sha256": "4a0a1f3a28c853c2d80092cabb1e1e09be0120a9dbd32d4cac72d37a12174f66", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/codex-inline-judge.json", "bytes": 4908, "sha256": "2b764b17563c4aefbcf52db9d7af8f095c2d2dbfb211655298f13e1231d334c6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/r3-nanbeige4.2-3b-q4-core-petshop/petshop/petshop-nanbeige4-2-3b-q4-2026-10-22T17-30-26-726Z-ag4n/postmortem.md", "bytes": 13242, "sha256": "3ad6bea927d185fa3cde70bb6aaf608d392bc73a51bf141e886818b758b64a65", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021.json", "bytes": 12605, "sha256": "9fdbf88b9b3e34ae71f9b18390dce08939284c1b618ad999a734f4765e611b24", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021-desktop.png", "bytes": 57814, "sha256": "32478b0929c528bdbc02b49247f3bcf6c5796903bac6219d291965846ef373b5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation7-2026-10-22/advisory-browser-evidence/3201d89021-mobile.png", "bytes": 56800, "sha256": "90defcc158dff2c59f29043fceef1cc1814759b6dbff8d57fb12ee9aa9f48572", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/workspace/tank-combat-arcade/index.html", "bytes": 12131, "sha256": "c84bdfb8b8d1ae4f7c50a0148306c4cd4000aa2cb00589103f15808f736172a4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-context.json", "bytes": 3309, "sha256": "33e97b04f6ee0f5544d215e6e2d428b0e63d7a2e936c8ab0e37adcf7abad5301", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-replay.json", "bytes": 1824, "sha256": "7926ca71ffa730cd47efa3e889e237cc88e46c37fc9de0f35fe259e65e37f8cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/workspace/tank-combat-arcade/index.html", "bytes": 12131, "sha256": "c84bdfb8b8d1ae4f7c50a0148306c4cd4000aa2cb00589103f15808f736172a4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-context.json", "bytes": 3309, "sha256": "33e97b04f6ee0f5544d215e6e2d428b0e63d7a2e936c8ab0e37adcf7abad5301", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-replay.json", "bytes": 1824, "sha256": "7926ca71ffa730cd47efa3e889e237cc88e46c37fc9de0f35fe259e65e37f8cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/facts.json", "bytes": 11961, "sha256": "9d6ca65412f192a188fdb58fc9648ae52e794e663c28b8943b2aa7e01683a940", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/result.json", "bytes": 945, "sha256": "69366f4d4a7f544bc152f63cd85a373e3f4c8000e33fee4fb7ff87ffc856d87f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/score.json", "bytes": 3208, "sha256": "e4d59f29e26d096e4df9d053af65d8e2ab7a4cce5c50d49454d8a0fba8f4d013", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/metrics.json", "bytes": 1315, "sha256": "021535c9e26559f56d78a2d6c28e79a588cff6f36cfd413f10947d6fcde5e6d2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/recording/chat-events.jsonl", "bytes": 5743508, "sha256": "5aef03b62e2b425b28e90334e5d4f54071c2a75d7f3bd76a93e9497ccd4f1b96", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/daemon.log", "bytes": 112926, "sha256": "6516900e66a633a1dc89dda62bb89fe41e05f02f5621b06a2e7979b0a57defc5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/result.json", "bytes": 945, "sha256": "69366f4d4a7f544bc152f63cd85a373e3f4c8000e33fee4fb7ff87ffc856d87f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/score.json", "bytes": 3208, "sha256": "e4d59f29e26d096e4df9d053af65d8e2ab7a4cce5c50d49454d8a0fba8f4d013", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/metrics.json", "bytes": 1315, "sha256": "021535c9e26559f56d78a2d6c28e79a588cff6f36cfd413f10947d6fcde5e6d2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/facts.json", "bytes": 11961, "sha256": "9d6ca65412f192a188fdb58fc9648ae52e794e663c28b8943b2aa7e01683a940", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/codex-inline-judge.json", "bytes": 4064, "sha256": "a458c92ae880e16eef7947cb81596ab3045343e31664158f0c0504fdbf804950", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-tankcombat/tankcombat/tankcombat-nanbeige4-2-3b-q4-2026-10-22T18-10-08-231Z-p7kb/postmortem.md", "bytes": 10852, "sha256": "c2aba6c4d411ee6650169065b1c76c7e09fd52a799376cb76bd44453d5fc73ff", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d.json", "bytes": 8574, "sha256": "5061be73daa05be172b3932999fe5d7e5d26b95137f5102c4d60e58a7ba17b77", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-start.png", "bytes": 37055, "sha256": "40a16dfc43ea5b2e66e5cf7e64ca0acab72957249357220011deead1df28d9b3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-desktop.png", "bytes": 441896, "sha256": "87021deca94e6c5ec888d8436f5b4c11f25caeb7025f92bb727f8ac3018c06ee", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-browser-evidence/6f24d6947d-mobile.png", "bytes": 127810, "sha256": "ae02720c6d03609db1998218c8a78b8ba5c6ba81d25c48fded1b71467d66b3fb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/types.ts", "bytes": 857, "sha256": "adefc2341bdf6daeb3bad9b802a539e188c0a5912412907c74f64162d5dca617", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/store.ts", "bytes": 1159, "sha256": "4d8f71bf3198a107666929a4fb7a21deacc2fe6cece809f5ba050c2f2244deab", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/migrate.ts", "bytes": 1212, "sha256": "ccc69cf66c64a43d9f254ad9235595a670bee9feef60cde723ed9ac141a362fa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/tests/migrate.test.ts", "bytes": 1239, "sha256": "4aa9e0093bfb73aca8768d371408d2b6858b7f7a35b88f99967afc31f9cb0c8a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/MIGRATION.md", "bytes": 1970, "sha256": "8cd648c39db296d0c8d4b33f30e045e7d167a3c32199eb50ffb8202ae22a2873", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e6f0947979-context.json", "bytes": 8717, "sha256": "8504f39e3fe553c2b8d2982d2936d727240f4f82e78209da60f84dc7be3e8531", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/types.ts", "bytes": 857, "sha256": "adefc2341bdf6daeb3bad9b802a539e188c0a5912412907c74f64162d5dca617", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/store.ts", "bytes": 1159, "sha256": "4d8f71bf3198a107666929a4fb7a21deacc2fe6cece809f5ba050c2f2244deab", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/handlers.ts", "bytes": 614, "sha256": "f151082616220d177a6e1bc1d525727b820376df24996db8e901924a3d2b1b85", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/src/migrate.ts", "bytes": 1212, "sha256": "ccc69cf66c64a43d9f254ad9235595a670bee9feef60cde723ed9ac141a362fa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/tests/migrate.test.ts", "bytes": 1239, "sha256": "4aa9e0093bfb73aca8768d371408d2b6858b7f7a35b88f99967afc31f9cb0c8a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/MIGRATION.md", "bytes": 1970, "sha256": "8cd648c39db296d0c8d4b33f30e045e7d167a3c32199eb50ffb8202ae22a2873", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/package.json", "bytes": 118, "sha256": "da7cfa9ba7722e4bdeb55cc9b373f3a311dfd6096b48587df9fe10286cd581fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/tsconfig.json", "bytes": 323, "sha256": "4d432e97e14419c8cd5438b66f9177e84275f76e029c7f13d30683d65fec7c38", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/workspace/user-store/README.md", "bytes": 233, "sha256": "e8ad3a935d40b5659206cfec22cc684ef10b2e5e673130d9e501d192d87d1194", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/schema-migration-seeded-source.json", "bytes": 2608, "sha256": "2f2ddc4464db935a92cb5775060c3c0e0a4b4ea8cfea2b74eedb671a0ef17fe3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e6f0947979-context.json", "bytes": 8717, "sha256": "8504f39e3fe553c2b8d2982d2936d727240f4f82e78209da60f84dc7be3e8531", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/facts.json", "bytes": 18181, "sha256": "4d45d31c8b152cdb71ee8811e35ccc27de00011bda88a5debe6c035680b3b4d6", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/result.json", "bytes": 1591, "sha256": "51de7310849d2389bb6d8e805ad40f9590d12c425731ee85e3360bd109fc699c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/score.json", "bytes": 5740, "sha256": "fb5fc43e36a146510b373ff0dab20bc5863c1152795486a0f6e422e300ae4375", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/metrics.json", "bytes": 2386, "sha256": "2c55ebe5f4165dd931e29afa0bc38f9a0112d4c4f0ee95b7318dfeeb0b180378", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/recording/chat-events.jsonl", "bytes": 2500669, "sha256": "c840963cc54024d01db42f9e2631851452be472072124da08534496f0ca724c2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/daemon.log", "bytes": 530324, "sha256": "7879d2f061d02f66f03c929fa10ebac459ed552360cc6e3780f3a29d5a2b7227", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/result.json", "bytes": 1591, "sha256": "51de7310849d2389bb6d8e805ad40f9590d12c425731ee85e3360bd109fc699c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/score.json", "bytes": 5740, "sha256": "fb5fc43e36a146510b373ff0dab20bc5863c1152795486a0f6e422e300ae4375", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/metrics.json", "bytes": 2386, "sha256": "2c55ebe5f4165dd931e29afa0bc38f9a0112d4c4f0ee95b7318dfeeb0b180378", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/facts.json", "bytes": 18181, "sha256": "4d45d31c8b152cdb71ee8811e35ccc27de00011bda88a5debe6c035680b3b4d6", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/codex-inline-judge.json", "bytes": 8744, "sha256": "6715a8c7922bff600c41cb6d05a7481d604a696d400dd628925a79ffe9a96838", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-schema-migration/schema-migration/schema-migration-nanbeige4-2-3b-q4-2026-10-22T18-22-33-585Z-muuf/postmortem.md", "bytes": 13597, "sha256": "f546c6877dc3834581ba4d71966a322f6c329b43e77be560afc9e1d170a24033", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e6f0947979.json", "bytes": 13979, "sha256": "05e1012fd96d773f3487f462f3f635d8b09a44183b94bc5714e34e954cb7805b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/src/machine.ts", "bytes": 1417, "sha256": "62edb9e9e61f55c399ba1d779eeba1c5e19a1b3a213bb1f7c87dcd181ea205fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/c77380beb8-context.json", "bytes": 6076, "sha256": "7b4fcb55e68a4d9242246cb35b0cef73c473cc501f95c9a10d0bdec9f8a2fe16", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/src/machine.ts", "bytes": 1417, "sha256": "62edb9e9e61f55c399ba1d779eeba1c5e19a1b3a213bb1f7c87dcd181ea205fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/tests/machine.test.ts", "bytes": 4926, "sha256": "12e2d36908178da95c12c9d5bf5d3bb7d8c67606f2ec4faa3ffd5fff62517997", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/package.json", "bytes": 205, "sha256": "6f3c6e522efed23d503e1270df18e6aef4f87a411c7be2d78cdb24842feb392d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/workspace/machine-kata/tsconfig.json", "bytes": 266, "sha256": "8bda3c0260964f6b5352b61969a681b7760822e954787a150c0e959839d389e3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/c77380beb8-context.json", "bytes": 6076, "sha256": "7b4fcb55e68a4d9242246cb35b0cef73c473cc501f95c9a10d0bdec9f8a2fe16", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/facts.json", "bytes": 14961, "sha256": "cd50df7b506e474e2587fb9abbaea0c3d2f9d87580c56a5f868dd3db3cf91cfa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/result.json", "bytes": 1920, "sha256": "6d68094d60d0b91cfe1684b3e9915a4c161c1c659a5fd9511d5943bdd92fdf01", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/score.json", "bytes": 5151, "sha256": "0d6788890e5bb88ce7c171aa76bf61594c1b8fc5cba002ba2d4f4e4d45f72297", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/metrics.json", "bytes": 2376, "sha256": "4e91872ad925fd90aace6c7c17615aaf0af78fc3e05323a5361a43ba655067b9", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/recording/chat-events.jsonl", "bytes": 628366, "sha256": "b8c7002117bcac17e4593b8ce429d300291eddbf56193bd5209a714a627a2c82", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/daemon.log", "bytes": 302083, "sha256": "15c429dc7d81cdffa57814f424a272ca403f98a66d35aca1f632ab8ff696001b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/result.json", "bytes": 1920, "sha256": "6d68094d60d0b91cfe1684b3e9915a4c161c1c659a5fd9511d5943bdd92fdf01", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/score.json", "bytes": 5151, "sha256": "0d6788890e5bb88ce7c171aa76bf61594c1b8fc5cba002ba2d4f4e4d45f72297", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/metrics.json", "bytes": 2376, "sha256": "4e91872ad925fd90aace6c7c17615aaf0af78fc3e05323a5361a43ba655067b9", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/facts.json", "bytes": 14961, "sha256": "cd50df7b506e474e2587fb9abbaea0c3d2f9d87580c56a5f868dd3db3cf91cfa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/codex-inline-judge.json", "bytes": 5326, "sha256": "53b6f3971ddbf63409934b49abfecdd96828da8a31d75c6307eae5a1a6b3e816", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-failing-tests-spec/failing-tests-spec/failing-tests-spec-nanbeige4-2-3b-q4-2026-10-22T18-33-35-458Z-5ddf/postmortem.md", "bytes": 12565, "sha256": "45b6c85c0838a63a0388885e7b3d9c1019bdd0363b31f5acebc184a7c9f940b1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/c77380beb8.json", "bytes": 47835, "sha256": "7e05821e82da9b0005ad1193e035886d6f872e2c64f23e32ae17a00b19d5687f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/workspace/pagination-service/package.json", "bytes": 120, "sha256": "4c5c2652bca3bf0d3a42c8997e49379234b83b6d089fd46bf4c4951fbec107e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/symptom-debug-brief-source.txt", "bytes": 1699, "sha256": "6acdb2ebc8cebb3d022d9de2c4dd7910aac473bf5fe8291b01b6b724fd0e7a09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e9987c93b8-context.json", "bytes": 983, "sha256": "5c7deb02479e0efd533a98e1a737b2471d5f5229968c28625c1f7d0752a43d47", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/workspace/pagination-service/lib/paginate.mjs", "bytes": 1189, "sha256": "23bc573089829f27703482c11dd75fd3ad8f2e826a2e9ebcd431a1f4a8ebd327", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/workspace/pagination-service/accept.mjs", "bytes": 1769, "sha256": "1a14b1c154b4faaca4d35d3d4124692e8bab41f43638161479c08a01fbf046a3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/symptom-debug-seeded-source.mjs", "bytes": 1193, "sha256": "48da8b5e9e8dbda4b437a74e0d3c4fe67e47b2c5b4ebf46d1aafe782b95a040f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/workspace/pagination-service/package.json", "bytes": 120, "sha256": "4c5c2652bca3bf0d3a42c8997e49379234b83b6d089fd46bf4c4951fbec107e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/symptom-debug-brief-source.txt", "bytes": 1699, "sha256": "6acdb2ebc8cebb3d022d9de2c4dd7910aac473bf5fe8291b01b6b724fd0e7a09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e9987c93b8-context.json", "bytes": 983, "sha256": "5c7deb02479e0efd533a98e1a737b2471d5f5229968c28625c1f7d0752a43d47", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/facts.json", "bytes": 10563, "sha256": "e638fe5b89cef30496681d186a52b88c28a6ab10e382ed4a928ecbbaab2b4db3", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/result.json", "bytes": 884, "sha256": "337533300afaf8dce4dba687504d65f88661348a33bc4df3ca67729da6f579f0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/score.json", "bytes": 2816, "sha256": "6902862bb0141482e11743f98c4f58512bad19a86d8bae77d70799e978c40ecf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/metrics.json", "bytes": 1308, "sha256": "db46bbd6f9580a48246523122345f85b285f618944b87b91c770b1052fe96213", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/recording/chat-events.jsonl", "bytes": 119050, "sha256": "7c58346cc5a89abd987df82560fe775ae0b22ac68b4fa2915683b64295d0e949", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/daemon.log", "bytes": 33661, "sha256": "dd8c4f2978313256aa41d4052979294b83a60ce0616ff31839b622a05b00bb72", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/result.json", "bytes": 884, "sha256": "337533300afaf8dce4dba687504d65f88661348a33bc4df3ca67729da6f579f0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/score.json", "bytes": 2816, "sha256": "6902862bb0141482e11743f98c4f58512bad19a86d8bae77d70799e978c40ecf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/metrics.json", "bytes": 1308, "sha256": "db46bbd6f9580a48246523122345f85b285f618944b87b91c770b1052fe96213", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/facts.json", "bytes": 10563, "sha256": "e638fe5b89cef30496681d186a52b88c28a6ab10e382ed4a928ecbbaab2b4db3", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/codex-inline-judge.json", "bytes": 5362, "sha256": "c402103f443f14061912ab7838beb3fb7636e980d8b424cb265246f645f1eb5a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-symptom-debug/symptom-debug/symptom-debug-nanbeige4-2-3b-q4-2026-10-22T18-36-23-202Z-b81v/postmortem.md", "bytes": 10170, "sha256": "4005ec6b6e2b34b9f0ac6cd775244776623bb3718ed7912119c96dc5fd85df1c", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-code-evidence/e9987c93b8.json", "bytes": 13303, "sha256": "a00d0ca4288b67e8ce4e5dec81e7f3e2edd70beaccf071a90981cd39133801d5", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2356, "sha256": "ffc59145a37be94387a0817b600ebc2a399b209913272329190c1cf3b11011ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2356, "sha256": "ffc59145a37be94387a0817b600ebc2a399b209913272329190c1cf3b11011ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-data-evidence/44098f1793-context.json", "bytes": 6809, "sha256": "f27ef26d52b956d40acb55fb9f9937a9e51acbaa08e97a6832a58b7af9407f56", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/customers_a.csv", "bytes": 1045, "sha256": "bc92617373571bdf0bef1c29ee05aad44a7a68274c0f7334b1ebe8aa5eedfe54", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/customers_b.csv", "bytes": 1087, "sha256": "ff167222077dd5a971ee88d6ab656d98b2b2051e84ceb223bcc893aab9186b09", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/data/raw/legacy_export.csv", "bytes": 1148, "sha256": "90e6abcfcf404de96b3ac6385e3ddc3265125552377f3c3428d5636d310f9afa", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/package.json", "bytes": 142, "sha256": "73ec6f76038de3a0aa11f4d469fcd61a74c6e8c501517f9f143cdfd3163750b5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tools/check_customers.mjs", "bytes": 1714, "sha256": "1b634245f2c014cb948067633bb88d751519519f1d2c5576e351ca69f8953eee", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tools/clean_customers.mjs", "bytes": 2356, "sha256": "ffc59145a37be94387a0817b600ebc2a399b209913272329190c1cf3b11011ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/workspace/customer-data-cleanup/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/facts.json", "bytes": 11955, "sha256": "a77e16f2c585b7d95643b97518f3df635ccbe0233aebf9ceef98adeec34f9216", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/result.json", "bytes": 1315, "sha256": "840bffc1b6243c6ea823f61c10e57ab6fdeb5d1371699a18f4dcfc5a8063517a", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/score.json", "bytes": 3269, "sha256": "91026565af0d4d89e074c3f5eabccb7e707744c1063c7339e014838dd3889cf2", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/metrics.json", "bytes": 2391, "sha256": "962a1d3c7af8e7ae499bede10c2e68f39b0baa0741969e92303458ad1b116985", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/recording/chat-events.jsonl", "bytes": 2188032, "sha256": "58fbec0663bed8813cf0b520efa6d3e7d4cdea1c3d29fafc6a558222a7dda8e4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/daemon.log", "bytes": 296364, "sha256": "eb13fb122ba8d2d583d651305ee62481fbb025ac20f28f789e63944afd49fa56", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/result.json", "bytes": 1315, "sha256": "840bffc1b6243c6ea823f61c10e57ab6fdeb5d1371699a18f4dcfc5a8063517a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/score.json", "bytes": 3269, "sha256": "91026565af0d4d89e074c3f5eabccb7e707744c1063c7339e014838dd3889cf2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/metrics.json", "bytes": 2391, "sha256": "962a1d3c7af8e7ae499bede10c2e68f39b0baa0741969e92303458ad1b116985", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/facts.json", "bytes": 11955, "sha256": "a77e16f2c585b7d95643b97518f3df635ccbe0233aebf9ceef98adeec34f9216", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/codex-inline-judge.json", "bytes": 7389, "sha256": "a187b9ddeaf19212ae3885ed551dded8b1e299d36883c6a8851fe20718872453", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/r3-nanbeige4.2-3b-q4-core-data-wrangle/data-wrangle/data-wrangle-nanbeige4-2-3b-q4-2026-10-22T18-38-27-865Z-1b78/postmortem.md", "bytes": 11432, "sha256": "f8e4f8dbddc241de86f653e3b58ae26c7ab66243d65546a9e59bda58d89f5ab0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation8-2026-10-22/advisory-data-evidence/44098f1793.json", "bytes": 8849, "sha256": "a78054d558347cc73a94d7940b0e742b656a98e9f16e279fc8ee6d8498223343", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/package.json", "bytes": 149, "sha256": "0aabbdefb60c2f087f78eedabdcf50516faad0a9e4a2b70929aeca9e4e8977cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/advisory-document-evidence/f5bf43f97a-diagnostics.json", "bytes": 5631, "sha256": "818426de165e65278fb62362294af3e459b870b9fee21fd19a076215f399f519", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/.gitignore", "bytes": 96, "sha256": "1eeb397af0f0360da802e8f7097ee5d84a00cba2b7637c2206ffd17f8b1192fc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/deploy.log", "bytes": 2199, "sha256": "57f36242602ccd9ebc7fd94b1ada97583733f14124fbb0fc120ce81c240b8983", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/hotfix.diff", "bytes": 493, "sha256": "0ed914f58cde932b0711e8b11d3f7b1c5cfd070deccc557e830f32b8796d51d0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/metrics.csv", "bytes": 967, "sha256": "d5bc0ad5881840027a211167b8c325b22a7743b42aacf64689fabb278752f8d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/oncall-chat.txt", "bytes": 1817, "sha256": "b91ef7fad9c8b2900d6785f8e1f23fb7687760543d38e7e084f133be649d49f4", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/evidence/timeline.md", "bytes": 1769, "sha256": "4dba0de1cf3eaa2bb3595f419411d03e7e08c4d75397de666cbe4960aec8a044", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/package.json", "bytes": 149, "sha256": "0aabbdefb60c2f087f78eedabdcf50516faad0a9e4a2b70929aeca9e4e8977cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/workspace/checkout-incident-postmortem/tsconfig.json", "bytes": 251, "sha256": "153ca5053dafbe39134da8fc461496a3e6c399f790f680b2313b9b0b0623a27b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/recording/chat-events.jsonl", "bytes": 340225, "sha256": "b429169916f866d7e1ed73613f5d06b326d744ddbb3da73b0b36df603c2c5bb5", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/facts.json", "bytes": 10446, "sha256": "773c850c1848d3e656b00a385f3927287bb87e26809fecf3dc26a6f64c5a5b01", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/daemon.log", "bytes": 134875, "sha256": "56ac4b93a39fce5c634ee626c71bcfebd49db64ba5e7edb270a9e3da50ce4909", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/log.txt", "bytes": 8393, "sha256": "f70add1e26b9cd0985332507115f2b9e20f74cf78b9ffb92dba57777dab20352", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/result.json", "bytes": 1357, "sha256": "a9d4e13381f53b8dc695f4065b2ffb8e59bfe706644f5cfa13f6aeaf46f74ccb", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/score.json", "bytes": 3127, "sha256": "e48d420504713c8b835b9c764e9a01fab13c0245f36dacf1f83d40670f2d64ec", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/metrics.json", "bytes": 1307, "sha256": "20670072eed15989f25e0eef4e23a13fdc3cbd1c399493f29da905e99ce9c9aa", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/facts.json", "bytes": 10446, "sha256": "773c850c1848d3e656b00a385f3927287bb87e26809fecf3dc26a6f64c5a5b01", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/codex-inline-judge.json", "bytes": 5845, "sha256": "0a218186bb4e95f84d94ff16f732d826c83c0d96ca2d72850699b1b48d6c4c75", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/r3-nanbeige4.2-3b-q4-core-incident-postmortem/incident-postmortem/incident-postmortem-nanbeige4-2-3b-q4-2026-10-22T19-10-46-947Z-8wog/postmortem.md", "bytes": 10464, "sha256": "01e3d56953a0b77c73f51dbcffdda1aaf7143525cb9c5c0d178da9c3449d8bd2", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation9-2026-10-22/advisory-document-evidence/f5bf43f97a.json", "bytes": 6733, "sha256": "1f1cc6f14a0ce4c978a14b8ca08205edaed4849068c36ad0b5f7ed3b1485c559", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/runlog.md", "bytes": 980, "sha256": "dc8c07a4b5b38faf80bb9c4a985f064a066b154c6be3a37c706498904d50f2c8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1252, "sha256": "7d2133cae2d1859c846760ad0082cb28d2df7483d9f7293be98f31d6fc1a33cc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/574fd8b47e-context.json", "bytes": 5943, "sha256": "ec968ff641f3561dab317edc5cf205302fe5bceda40f9ad1560747ca9801e18d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/runlog.md", "bytes": 980, "sha256": "dc8c07a4b5b38faf80bb9c4a985f064a066b154c6be3a37c706498904d50f2c8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/halt-report.md", "bytes": 1252, "sha256": "7d2133cae2d1859c846760ad0082cb28d2df7483d9f7293be98f31d6fc1a33cc", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/runbook.md", "bytes": 1200, "sha256": "f23cd39c4c8cdf3f4e127f6e73a3bc7d9f6ed1ac7f21427731fe4a53bf51066f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/services.json", "bytes": 209, "sha256": "26de7023ce8cbbc14edc0379ed9264d71589f5fd3ed8ad5710e00c7cd1c4db98", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/manifest-checksum.txt", "bytes": 15, "sha256": "6926c0a84ff11970044b488340865f7fff979810d12ad66f808f621ab02df9a1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/config.json", "bytes": 59, "sha256": "82765fbb717cd140f69b49de361826d5aa9ae66136cd1f8a0d8ce029c05891fd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/state/backup.json", "bytes": 42, "sha256": "d9c3369c7b57500dae64bf9bc844ed98561f3d2d5a0a55a512bdbb11744fc38e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/checks/expected-counts.md", "bytes": 69, "sha256": "0230e59308c021cc9085ecc73b0dcc4dd2a79e5ff631667ddac29172ac4543f1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/workspace/quarterly-maintenance-window/maintenance/README.md", "bytes": 85, "sha256": "83ad1ce6bd9fefcf24d982fe55a543282f769e8bf23e0c0d20654cd2d6bb4c0e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/574fd8b47e-context.json", "bytes": 5943, "sha256": "ec968ff641f3561dab317edc5cf205302fe5bceda40f9ad1560747ca9801e18d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/facts.json", "bytes": 13198, "sha256": "12f0ea68f3916f4a9bc1a1ba1af413ebecaa0356a8956336ac69e61e890ccadf", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/result.json", "bytes": 1017, "sha256": "ae7f412072188991921b8f471f46a2de9cfc6fe67ec0049d07f7c3be1dcf832b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/score.json", "bytes": 2991, "sha256": "aea2abb23e5a7572ded59f78732d9ffaf5ea7cb1419a5e68592bb501234c6bfd", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/metrics.json", "bytes": 2377, "sha256": "f2e41555f60281164dfd5a24398f4a2fdc9fbdb0605da22af2fb74e12e41e65d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/recording/chat-events.jsonl", "bytes": 1041906, "sha256": "0c51281becfa8f4598b9f67d977b0e45bf5817d67497574d93cc1be8814b974c", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/daemon.log", "bytes": 108214, "sha256": "9813dc803737a00c274be1d23e861d11c7cc52b2bd2991fe9e03f08e494ac255", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/result.json", "bytes": 1017, "sha256": "ae7f412072188991921b8f471f46a2de9cfc6fe67ec0049d07f7c3be1dcf832b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/score.json", "bytes": 2991, "sha256": "aea2abb23e5a7572ded59f78732d9ffaf5ea7cb1419a5e68592bb501234c6bfd", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/metrics.json", "bytes": 2377, "sha256": "f2e41555f60281164dfd5a24398f4a2fdc9fbdb0605da22af2fb74e12e41e65d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/facts.json", "bytes": 13198, "sha256": "12f0ea68f3916f4a9bc1a1ba1af413ebecaa0356a8956336ac69e61e890ccadf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/codex-inline-judge.json", "bytes": 8967, "sha256": "e62dc2fbda516372d8472df59f0f7e5c352ace30bc8e32310e542be3a955bf02", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-ops-runbook-anomaly/ops-runbook-anomaly/ops-runbook-anomaly-nanbeige4-2-3b-q4-2026-10-22T20-05-57-936Z-geid/postmortem.md", "bytes": 10221, "sha256": "06f5027081cf29a33417cfea90bcb1133092a2aeb3258797d9bc3314ba8f970a", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/574fd8b47e.json", "bytes": 16125, "sha256": "5f56658ba249c1b2799c08723e740b46bb6a9d404c5a7a0ef00dd0b0881c8057", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/workspace/harbourview-office-relocation/plan.md", "bytes": 4224, "sha256": "5dc690da8245200c50745a6d4fc806f5147344b2b96a23305d8f93b6773bd6cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/558843afd6-context.json", "bytes": 3077, "sha256": "2812ac2ad45c5260ed231fc5d4351b7e2a825a5c23c0683bf0318be3ae6155ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/workspace/harbourview-office-relocation/plan.md", "bytes": 4224, "sha256": "5dc690da8245200c50745a6d4fc806f5147344b2b96a23305d8f93b6773bd6cb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/workspace/harbourview-office-relocation/brief.md", "bytes": 732, "sha256": "c477d71f7e43fdbca2ee8909ce78c886eba0c6cf3b431ff9358284c53031b34d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/workspace/harbourview-office-relocation/team.md", "bytes": 150, "sha256": "1fddabd1da50c41728b8f834ba787d5469a557e7faabb1efce8c3ae0d3c90912", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/558843afd6-context.json", "bytes": 3077, "sha256": "2812ac2ad45c5260ed231fc5d4351b7e2a825a5c23c0683bf0318be3ae6155ed", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/facts.json", "bytes": 9969, "sha256": "86341c4abb7f4c6ea06bc47683e5f86388f75d80c982ac77cc526951456aafd0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/result.json", "bytes": 925, "sha256": "2fe9e878e1317a34acd372325fc81dc9cbdc81c7f3bcbc43525f6cc1d047b7af", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/score.json", "bytes": 2910, "sha256": "39f9327c6052e134f4a3e68f0d916fc9d86b62fc0a22b1a26c0dd7de582dcdbe", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/metrics.json", "bytes": 1308, "sha256": "2929ff7fd3c1490e35338e7c207f385a37e38209d2aae89e4e4337bfba6dbd07", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/recording/chat-events.jsonl", "bytes": 549550, "sha256": "52eaa7cffc94a79a844ba0c3ee8be10f9b9dcfaf6415a6bac6c52a3edd841d30", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/daemon.log", "bytes": 29221, "sha256": "0945e4d28b88b4b772d57857ba9e9885624a5f685f99269f6af43f15dc40c8b8", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/result.json", "bytes": 925, "sha256": "2fe9e878e1317a34acd372325fc81dc9cbdc81c7f3bcbc43525f6cc1d047b7af", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/score.json", "bytes": 2910, "sha256": "39f9327c6052e134f4a3e68f0d916fc9d86b62fc0a22b1a26c0dd7de582dcdbe", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/metrics.json", "bytes": 1308, "sha256": "2929ff7fd3c1490e35338e7c207f385a37e38209d2aae89e4e4337bfba6dbd07", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/facts.json", "bytes": 9969, "sha256": "86341c4abb7f4c6ea06bc47683e5f86388f75d80c982ac77cc526951456aafd0", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/codex-inline-judge.json", "bytes": 4802, "sha256": "b50afe8015589ef13e9ddc0281e43e748bfff5bdee15c637974a62b773b14d6f", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-plan-and-estimate/plan-and-estimate/plan-and-estimate-nanbeige4-2-3b-q4-2026-10-22T20-10-49-585Z-mk84/postmortem.md", "bytes": 10308, "sha256": "d41ab3143035ad3a1f761d147f69b46872e7d0d1c26cefa460b56eb694f0eebf", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/558843afd6.json", "bytes": 8453, "sha256": "246b72d0cd409033c628aa08d6cede3150a35d97a7d087e8431fa2d4dacd4e23", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 9168, "sha256": "125a03aefebf83f65c9aaad3cab81886a0e6d3716c6884aedf8620a59878f69b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/b317f0577b-context.json", "bytes": 4744, "sha256": "651b0c7145146136e1af8953fbce55091937e158a417fff8612dc146e4c8c4e0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/synthesis.md", "bytes": 9168, "sha256": "125a03aefebf83f65c9aaad3cab81886a0e6d3716c6884aedf8620a59878f69b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/memo-product.md", "bytes": 216, "sha256": "6c027b5d65b0300d871c9b5c4b73b38d0569e58a381480ebf681888f28b53e2e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/memo-engineering.md", "bytes": 256, "sha256": "77944097ac4fb8313f3f00290f85d9cb12b34c1c340a06dd67644d3fef3e0077", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/finance.csv", "bytes": 93, "sha256": "e24b11e1c2faa0a515ab40967299793a1e0f8de591f8135c98b8f200daa028d1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/org.md", "bytes": 111, "sha256": "8b0fd735b61fb96c24801f009fc28592c9aca318c79cc82fe9d1a6f1b1cfd6fb", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/memo-oldplan.md", "bytes": 156, "sha256": "ec40515a54b19eae66fd83a0f5557837c439f10ee24e5a9f3da4b2430860d68f", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/workspace/skylark-launch-synthesis/style.md", "bytes": 97, "sha256": "729e515322284b7614b2439c4a2ac7a5cf820997b91979323cd62a34910d370d", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/b317f0577b-context.json", "bytes": 4744, "sha256": "651b0c7145146136e1af8953fbce55091937e158a417fff8612dc146e4c8c4e0", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/facts.json", "bytes": 17041, "sha256": "1b8fe000c83564138e5071b29366186ac8cc312455c917bf2178d0927e4f7854", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/result.json", "bytes": 2335, "sha256": "8ea9b39fa3b269a5d149c0be87170740a06b668b9f069efc26ea41365b5f7a9e", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/score.json", "bytes": 4562, "sha256": "955e1e917fb61eaf406fdbf99917bee1d5fc4840c3049ccdb3a7d13132f2c26b", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/metrics.json", "bytes": 2386, "sha256": "cee35f8025838fe145131a47e846e43ed7b22a382fd164612578c50dd10078e1", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/recording/chat-events.jsonl", "bytes": 3908916, "sha256": "368e5c6d29dc86d05ae06c791b3778f6a6e6cc980af0fd7c2bcaee42b2b96f87", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/daemon.log", "bytes": 234610, "sha256": "b9d1623fd92beefca72cdff80f5d586b3b5fdbb5d478e3761d4a46caa888f0ea", "verifiedAgainstRecorded": true }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/result.json", "bytes": 2335, "sha256": "8ea9b39fa3b269a5d149c0be87170740a06b668b9f069efc26ea41365b5f7a9e", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/score.json", "bytes": 4562, "sha256": "955e1e917fb61eaf406fdbf99917bee1d5fc4840c3049ccdb3a7d13132f2c26b", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/metrics.json", "bytes": 2386, "sha256": "cee35f8025838fe145131a47e846e43ed7b22a382fd164612578c50dd10078e1", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/facts.json", "bytes": 17041, "sha256": "1b8fe000c83564138e5071b29366186ac8cc312455c917bf2178d0927e4f7854", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/codex-inline-judge.json", "bytes": 7610, "sha256": "85f07f87aa540b60e79e5463cd774aa076cbd87e44ebe8e6f8bd5f79ce58629d", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/r3-nanbeige4.2-3b-q4-core-conflict-synthesis/conflict-synthesis/conflict-synthesis-nanbeige4-2-3b-q4-2026-10-22T20-13-26-474Z-j1fn/postmortem.md", "bytes": 12036, "sha256": "2cadc85b92c279e14710afc0a65dddbc4fd39cf331709fb014f5d90ba295b587", "verifiedAgainstRecorded": false }, { "path": "evals/runs/nanbeige-32k-core-continuation10-2026-10-22/advisory-document-evidence/b317f0577b.json", "bytes": 13126, "sha256": "7335713a34d08aefed5f436efad9d682a01f2d6f6eb158c6a209ef7563447c07", "verifiedAgainstRecorded": false } ] }