Files
cdsl-cad/backend/live-evals/20260831T095431Z/report.json
T
2026-09-01 14:10:23 +08:00

846 lines
29 KiB
JSON

{
"status": "failed",
"author": {
"provider": "openai",
"model": "gpt-5.4-mini"
},
"reviewer": {
"provider": "openai",
"model": "gpt-5.5"
},
"author_capability": {
"schema_hash": "fb899a5de2f869b648df88a32d04a5eee92ae3d193ab23ee641892d6b59cd8e7",
"tool_count": 9,
"supported": true,
"failures": [],
"usage": {
"prompt_tokens": 41789,
"completion_tokens": 1472
},
"mode": "openai_compatible_tool_calling"
},
"reviewer_capability": {
"schema_hash": "fb899a5de2f869b648df88a32d04a5eee92ae3d193ab23ee641892d6b59cd8e7",
"tool_count": 9,
"supported": true,
"failures": [],
"usage": {
"prompt_tokens": 41789,
"completion_tokens": 536
},
"mode": "openai_compatible_tool_calling"
},
"results": [
{
"scenario": "rectangular_plate",
"repetition": 1,
"success": true,
"started_at": "2026-08-31T09:55:55.322896+00:00",
"finished_at": "2026-08-31T09:56:56.059317+00:00",
"terminal": {
"taskId": "cad_f18ebbd436c7",
"lifecycle": "completed",
"revisionId": "rev_001"
},
"projection": {
"schema_version": "3.0",
"task_id": "cad_f18ebbd436c7",
"phase": "COMPLETED",
"lifecycle": "completed",
"state_version": 9,
"active_revision": "rev_001",
"current_revision": "rev_001",
"published_revision": "rev_001",
"pending_action": null,
"active_candidate_id": "",
"repair_required": false,
"last_error": "",
"action_ledger_summary": [
{
"sequence": 1,
"at": "2026-08-31 09:56:04",
"draft_revision": 1,
"event": "requirements_draft_updated",
"item_count": 1
},
{
"sequence": 2,
"at": "2026-08-31 09:56:08",
"draft_revision": 1,
"event": "requirements_review_requested"
},
{
"sequence": 3,
"at": "2026-08-31 09:56:14",
"contract_hash": "5a77d1672adaecd4a1551764149b95abce0fdc2ed5d434b460e93bc5085caee9",
"event": "requirements_contract_frozen",
"requirement_count": 1
},
{
"sequence": 4,
"at": "2026-08-31 09:56:26",
"action_id": "act_004",
"atomic_id": "extrude_add_blind",
"contract_hash": "fa9ffc9fd2141453fa18a389813dde2e2ef3fb4c7ef5d05e61d51d614309e95f",
"event": "proposed",
"expected_change": "Create a single solid rectangular plate measuring 80 mm by 50 mm by 8 mm in the current model.",
"intent": "Build the base plate body to satisfy the dimensional requirement.",
"requirement_ids": [
"req_001"
],
"working_head": "cad_f18ebbd436c7:root:v3"
},
{
"sequence": 5,
"at": "2026-08-31 09:56:34",
"action_id": "act_004",
"candidate_id": "candidate_fd1dd253fd90e2f2",
"event": "candidate_building",
"fragment_hash": "01a2f45e7ae8ef9e48f5dd39c74d65f0e83c82cf0c234deada7edd1507496fae"
},
{
"sequence": 6,
"at": "2026-08-31 09:56:35",
"action_id": "act_004",
"candidate_id": "candidate_fd1dd253fd90e2f2",
"claim_results": [
{
"claim_id": "claim_001",
"claim_kind": "solid_count_equals",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_002",
"claim_kind": "single_connected_body",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_003",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 80.0,
"axis": "x",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_004",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 50.0,
"axis": "y",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_005",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 8.0,
"axis": "z",
"tolerance_mm": 0.0
},
"status": "pass"
}
],
"event": "candidate_built"
},
{
"sequence": 7,
"at": "2026-08-31 09:56:44",
"action_id": "act_004",
"actual_atomic_id": "extrude_add_blind",
"candidate_id": "candidate_fd1dd253fd90e2f2",
"coverage": [
{
"claim_id": "claim_001",
"claim_kind": "solid_count_equals",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_002",
"claim_kind": "single_connected_body",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_003",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 80.0,
"axis": "x",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_004",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 50.0,
"axis": "y",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_005",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 8.0,
"axis": "z",
"tolerance_mm": 0.0
},
"status": "pass"
}
],
"event": "accepted",
"fragment_hash": "01a2f45e7ae8ef9e48f5dd39c74d65f0e83c82cf0c234deada7edd1507496fae",
"review_path": "revisions/rev_001/candidate-review.json",
"revision_id": "rev_001",
"selector_snapshot_id": "",
"working_head_before": "cad_f18ebbd436c7:root:v3"
},
{
"sequence": 8,
"at": "2026-08-31 09:56:47",
"claim_results": [
{
"claim_id": "claim_001",
"claim_kind": "solid_count_equals",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_002",
"claim_kind": "single_connected_body",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_003",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 80.0,
"axis": "x",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_004",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 50.0,
"axis": "y",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_005",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 8.0,
"axis": "z",
"tolerance_mm": 0.0
},
"status": "pass"
}
],
"event": "final_validation_requested",
"revision_id": "rev_001"
},
{
"sequence": 9,
"at": "2026-08-31 09:56:56",
"claim_results": [
{
"claim_id": "claim_001",
"claim_kind": "solid_count_equals",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_002",
"claim_kind": "single_connected_body",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_003",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 80.0,
"axis": "x",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_004",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 50.0,
"axis": "y",
"tolerance_mm": 0.0
},
"status": "pass"
},
{
"claim_id": "claim_005",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 8.0,
"axis": "z",
"tolerance_mm": 0.0
},
"status": "pass"
}
],
"event": "completed",
"final_review_path": "revisions/rev_001/final-review.json",
"revision_id": "rev_001"
}
],
"revisions": [
{
"revision_id": "rev_001",
"status": "success",
"visibility": "final",
"cdsl_path": "revisions/rev_001/model.cdsl.json",
"step_path": "revisions/rev_001/model.step",
"glb_path": "revisions/rev_001/model.glb",
"report_path": "revisions/rev_001/rebuild-report.json",
"candidate_review_path": "revisions/rev_001/candidate-review.json"
}
]
},
"usage": {
"calls": 8,
"prompt_tokens": 66683,
"completion_tokens": 2143,
"context_chars": 43886,
"records": [
{
"completion_tokens": 448,
"context_chars": 878,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 5234,
"provider_id": "openai",
"raw_arguments_hash": "004b933bbdcee2ec8a612472e0f5c36813bb55a2624b0100429390ac61b8076e",
"tool": "submit_requirements_draft_batch",
"total_tokens": 5682
},
{
"completion_tokens": 79,
"context_chars": 1673,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 5448,
"provider_id": "openai",
"raw_arguments_hash": "44136fa355b3678a1146ad16f7e8649e94fb4fc21fe77e8310c060f61caaff8a",
"tool": "finalize_requirements_draft",
"total_tokens": 5527
},
{
"completion_tokens": 71,
"context_chars": 1159,
"model_id": "gpt-5.5",
"prompt_tokens": 5011,
"provider_id": "openai",
"raw_arguments_hash": "00996cae282886750e3169d668db3145d5aad3fa73c825b1b9d5d423613bdb52",
"role": "reviewer",
"tool": "review_requirements",
"total_tokens": 5082
},
{
"completion_tokens": 569,
"context_chars": 1281,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 4960,
"provider_id": "openai",
"raw_arguments_hash": "4bff6094579f7f30f0f888d2044ed6ad56b24ffc5117b4e25e1af85714df8bb7",
"tool": "propose_next_action",
"total_tokens": 5529
},
{
"completion_tokens": 402,
"context_chars": 2239,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 5634,
"provider_id": "openai",
"raw_arguments_hash": "75e87f0b7034b3d9982873a523c40ff81dc255a14b2dbf23a1bf63d06be85113",
"tool": "submit_cdsl_fragment",
"total_tokens": 6036
},
{
"completion_tokens": 277,
"context_chars": 21836,
"model_id": "gpt-5.5",
"prompt_tokens": 19246,
"provider_id": "openai",
"raw_arguments_hash": "dc76df5308ce57cdbdd3c28c7f43d323996736ba584d9b7f0b430b3f17f701ad",
"role": "reviewer",
"tool": "review_candidate",
"total_tokens": 19523
},
{
"completion_tokens": 65,
"context_chars": 4046,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 6074,
"provider_id": "openai",
"raw_arguments_hash": "44136fa355b3678a1146ad16f7e8649e94fb4fc21fe77e8310c060f61caaff8a",
"tool": "complete_task",
"total_tokens": 6139
},
{
"completion_tokens": 232,
"context_chars": 10774,
"model_id": "gpt-5.5",
"prompt_tokens": 15076,
"provider_id": "openai",
"raw_arguments_hash": "65cb6d82f5b65bb4c0e32cea6db1c093949c76b5cdc6b1871cf772934a7915eb",
"role": "reviewer",
"tool": "review_final",
"total_tokens": 15308
}
]
},
"checks": {
"expected_terminal_phase": true,
"token_budget": true,
"call_budget": true,
"author_turn_budget": true,
"reviewer_turn_budget": true,
"required_claims": true,
"required_operations": true,
"required_operation_alternative": true,
"immutable_artifacts": true,
"unique_revisions": true,
"raw_argument_audit": true,
"no_schema_or_decision_rejections": true
},
"rejection_codes": [],
"artifact_root": "/Users/lk/Projects/cdsl-cad/backend/live-evals/20260831T095431Z/artifacts/cad_f18ebbd436c7"
},
{
"scenario": "simple_flange",
"repetition": 1,
"success": false,
"started_at": "2026-08-31T09:56:56.061618+00:00",
"finished_at": "2026-08-31T10:00:03.260135+00:00",
"terminal": {
"taskId": "cad_e1261e227f5c",
"lifecycle": "failed",
"code": "FAILED_AUTHOR_FORMAT",
"message": "Author repeatedly failed the same canonical schema or state contract."
},
"projection": {
"schema_version": "3.0",
"task_id": "cad_e1261e227f5c",
"phase": "FAILED",
"lifecycle": "failed",
"state_version": 9,
"active_revision": "rev_001",
"current_revision": "rev_001",
"published_revision": "",
"pending_action": null,
"active_candidate_id": "",
"repair_required": false,
"last_error": "FAILED_AUTHOR_FORMAT",
"action_ledger_summary": [
{
"sequence": 10,
"at": "2026-08-31 09:57:11",
"draft_revision": 1,
"event": "requirements_draft_updated",
"item_count": 1
},
{
"sequence": 11,
"at": "2026-08-31 09:57:15",
"draft_revision": 1,
"event": "requirements_review_requested"
},
{
"sequence": 12,
"at": "2026-08-31 09:57:23",
"contract_hash": "7a31c9479ee38b16ce3aa77a1eece02a1a889b8b98cba8d11059977acb26118a",
"event": "requirements_contract_frozen",
"requirement_count": 1
},
{
"sequence": 13,
"at": "2026-08-31 09:57:34",
"action_id": "act_004",
"atomic_id": "extrude_add_blind",
"contract_hash": "fa9ffc9fd2141453fa18a389813dde2e2ef3fb4c7ef5d05e61d51d614309e95f",
"event": "proposed",
"expected_change": "Create a 120 mm outer-diameter, 12 mm-thick cylindrical base solid in millimetres.",
"intent": "Establish the flange body before adding the centered 40 mm through bore.",
"requirement_ids": [
"req_001"
],
"working_head": "cad_e1261e227f5c:root:v3"
},
{
"sequence": 14,
"at": "2026-08-31 09:57:44",
"action_id": "act_004",
"candidate_id": "candidate_585ceff4e4616a5c",
"event": "candidate_building",
"fragment_hash": "95bd1e739aed6d4a791a162b2ce05d84d0a52f4d18dc87c11dc7d6e8d7761078"
},
{
"sequence": 15,
"at": "2026-08-31 09:57:46",
"action_id": "act_004",
"candidate_id": "candidate_585ceff4e4616a5c",
"claim_results": [
{
"claim_id": "claim_001",
"claim_kind": "solid_count_equals",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_002",
"claim_kind": "single_connected_body",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_003",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 120.0,
"axis": "x",
"tolerance_mm": 0.1
},
"status": "pass"
},
{
"claim_id": "claim_004",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 120.0,
"axis": "y",
"tolerance_mm": 0.1
},
"status": "pass"
},
{
"claim_id": "claim_005",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 12.0,
"axis": "z",
"tolerance_mm": 0.1
},
"status": "pass"
},
{
"claim_id": "claim_006",
"claim_kind": "through_cylindrical_bore",
"deterministic": true,
"evidence": {
"reason": "the target bore has not been introduced at this checkpoint"
},
"status": "pending"
}
],
"event": "candidate_built"
},
{
"sequence": 16,
"at": "2026-08-31 09:57:57",
"action_id": "act_004",
"actual_atomic_id": "extrude_add_blind",
"candidate_id": "candidate_585ceff4e4616a5c",
"coverage": [
{
"claim_id": "claim_001",
"claim_kind": "solid_count_equals",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_002",
"claim_kind": "single_connected_body",
"deterministic": true,
"evidence": {
"solid_count": 1
},
"status": "pass"
},
{
"claim_id": "claim_003",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 120.0,
"axis": "x",
"tolerance_mm": 0.1
},
"status": "pass"
},
{
"claim_id": "claim_004",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 120.0,
"axis": "y",
"tolerance_mm": 0.1
},
"status": "pass"
},
{
"claim_id": "claim_005",
"claim_kind": "bbox_dimension_mm",
"deterministic": true,
"evidence": {
"actual_mm": 12.0,
"axis": "z",
"tolerance_mm": 0.1
},
"status": "pass"
},
{
"claim_id": "claim_006",
"claim_kind": "through_cylindrical_bore",
"deterministic": true,
"evidence": {
"reason": "the target bore has not been introduced at this checkpoint"
},
"status": "pending"
}
],
"event": "accepted",
"fragment_hash": "95bd1e739aed6d4a791a162b2ce05d84d0a52f4d18dc87c11dc7d6e8d7761078",
"review_path": "revisions/rev_001/candidate-review.json",
"revision_id": "rev_001",
"selector_snapshot_id": "",
"working_head_before": "cad_e1261e227f5c:root:v3"
},
{
"sequence": 17,
"at": "2026-08-31 09:58:05",
"action_id": "act_008",
"atomic_id": "hole_blind",
"contract_hash": "f62e12603feeca6d7e83acdab651fcb023e08a46c39fd1509e9dcb81d4f0c721",
"event": "proposed",
"expected_change": "Add a centered 40 mm diameter blind hole through the 12 mm flange thickness, removing material from one face to create the through bore.",
"intent": "Introduce the centered bore required by the flange specification while keeping the body a single connected solid.",
"requirement_ids": [
"req_001"
],
"working_head": "cad_e1261e227f5c:rev_001:v7"
},
{
"sequence": 18,
"at": "2026-08-31 10:00:03",
"event": "failed_author_format",
"field_errors": [],
"tool": "submit_cdsl_fragment"
}
],
"revisions": [
{
"revision_id": "rev_001",
"status": "success",
"visibility": "checkpoint",
"cdsl_path": "revisions/rev_001/model.cdsl.json",
"step_path": "revisions/rev_001/model.step",
"glb_path": "revisions/rev_001/model.glb",
"report_path": "revisions/rev_001/rebuild-report.json",
"candidate_review_path": "revisions/rev_001/candidate-review.json"
}
]
},
"usage": {
"calls": 10,
"prompt_tokens": 72681,
"completion_tokens": 10235,
"context_chars": 50920,
"records": [
{
"completion_tokens": 548,
"context_chars": 934,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 5242,
"provider_id": "openai",
"raw_arguments_hash": "0624ba33ca1696c191aaa3d60abfdeb4ef805d6ec4f9a949d70001dcebf3d978",
"tool": "submit_requirements_draft_batch",
"total_tokens": 5790
},
{
"completion_tokens": 140,
"context_chars": 1931,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 5510,
"provider_id": "openai",
"raw_arguments_hash": "44136fa355b3678a1146ad16f7e8649e94fb4fc21fe77e8310c060f61caaff8a",
"tool": "finalize_requirements_draft",
"total_tokens": 5650
},
{
"completion_tokens": 71,
"context_chars": 1377,
"model_id": "gpt-5.5",
"prompt_tokens": 5069,
"provider_id": "openai",
"raw_arguments_hash": "00996cae282886750e3169d668db3145d5aad3fa73c825b1b9d5d423613bdb52",
"role": "reviewer",
"tool": "review_requirements",
"total_tokens": 5140
},
{
"completion_tokens": 566,
"context_chars": 1347,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 4980,
"provider_id": "openai",
"raw_arguments_hash": "7e4557b99bdcb3a86e6f5179d210a593366a7a34ba331216cc1a6025bf6f438b",
"tool": "propose_next_action",
"total_tokens": 5546
},
{
"completion_tokens": 529,
"context_chars": 2296,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 5661,
"provider_id": "openai",
"raw_arguments_hash": "23829cf9a299ebb4730966c58369597723709d4bbc0ebc4984898749ab2c97f1",
"tool": "submit_cdsl_fragment",
"total_tokens": 6190
},
{
"completion_tokens": 269,
"context_chars": 22280,
"model_id": "gpt-5.5",
"prompt_tokens": 19510,
"provider_id": "openai",
"raw_arguments_hash": "bc2d6c0dd17595091d5c45a76988fb8911228fc637435ec306df256aed0de72c",
"role": "reviewer",
"tool": "review_candidate",
"total_tokens": 19779
},
{
"completion_tokens": 375,
"context_chars": 4552,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 6198,
"provider_id": "openai",
"raw_arguments_hash": "013a8a345baf8848619ae741c8375c03fab94771b062ce641d5c46f2ddfb0e7e",
"tool": "propose_next_action",
"total_tokens": 6573
},
{
"completion_tokens": 3431,
"context_chars": 5167,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 6787,
"provider_id": "openai",
"raw_arguments_hash": "65863e9517d4417def1a18f5cf89c0f77ffe21f82e4262b63340c81aec568819",
"tool": "submit_cdsl_fragment",
"total_tokens": 10218
},
{
"completion_tokens": 3690,
"context_chars": 5518,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 6862,
"provider_id": "openai",
"raw_arguments_hash": "d56f3f8cfec1a7e5619c9a001110cbac4d260b6a3ce226257f5968a66b48619c",
"tool": "submit_cdsl_fragment",
"total_tokens": 10552
},
{
"completion_tokens": 616,
"context_chars": 5518,
"model_id": "gpt-5.4-mini",
"prompt_tokens": 6862,
"provider_id": "openai",
"raw_arguments_hash": "d287cddf623758d4f47e68663ba7cd1aef15a298c23221f5dbcf22959676e0df",
"tool": "submit_cdsl_fragment",
"total_tokens": 7478
}
]
},
"checks": {
"expected_terminal_phase": false,
"token_budget": true,
"call_budget": true,
"author_turn_budget": true,
"reviewer_turn_budget": true,
"required_claims": true,
"required_operations": true,
"required_operation_alternative": false,
"immutable_artifacts": true,
"unique_revisions": true,
"raw_argument_audit": true,
"no_schema_or_decision_rejections": true
},
"rejection_codes": [],
"artifact_root": "/Users/lk/Projects/cdsl-cad/backend/live-evals/20260831T095431Z/artifacts/cad_e1261e227f5c"
}
],
"suite": "smoke",
"require_live": true,
"report_root": "/Users/lk/Projects/cdsl-cad/backend/live-evals/20260831T095431Z"
}