task_detail
validation/gpt-5-4/flux-pr-891/task_detail.v1.json
16931 bytes
{
"version": "v1",
"materialization": {
"mode": "canonical"
},
"identity": {
"variant_id": "gpt-5.4",
"requested_model": "gpt-5.4",
"model_key": "gpt-5-4",
"run_id": "2026-02-28__13-38-10__gpt-5-4",
"task_id": "flux-pr-891",
"trial_name": "flux-pr-891.1-of-1.2026-02-28__13-38-10__gpt-5-4"
},
"decision": {
"status": "pass",
"primary_signal": "tests",
"matrix_status": "pass_with_warn",
"reason_codes": [
"primary_tests",
"tests_passed",
"equivalence_non_equivalent",
"decision_conflict",
"review_fail"
],
"flags": [
"equiv_warn",
"review_warn"
],
"lane_report": {
"lane": "lane_unknown",
"source": "lane_unknown",
"reasons": [
"lane_unknown"
]
},
"publish": {
"include_in_leaderboard": true,
"weak_signal_risk": false
},
"rescue": {
"candidate": false,
"eligible": false,
"decision": "not_candidate",
"required_artifacts_ok": true
}
},
"tests": {
"outcome": "pass",
"strength": "strong",
"commands": [
"go test -C v2 ./... -count=1 -timeout=300s"
],
"command_results": [
{
"command": "go test -C v2 ./... -count=1 -timeout=300s",
"gold_outcome": "pass",
"gold_tb_trial_name": "flux-pr-891.1-of-1.validate-gold-1772831021762987000",
"gold_failure_mode": "unset",
"agent_outcome": "pass",
"agent_tb_trial_name": "flux-pr-891.1-of-1.validate-override-1772831156280377000",
"agent_failure_mode": "unset",
"agent_retried": false
}
],
"partial_score": {
"score": 1,
"numerator": 1,
"denominator": 1,
"level": "command",
"provenance": "fallback_command_level",
"reason": "test_case_detail_unavailable"
},
"tb": {
"trial_name": "flux-pr-891.1-of-1.2026-02-28__13-38-10__gpt-5-4",
"is_resolved": true,
"failure_mode": "unset",
"duration_sec": 247.624359125,
"time_budget_sec": 1200
},
"cache": {
"cache_hit": false,
"cache_miss_reason": "missing_pinned_dependencies",
"environment_group_id": "eg_46b6a6c82a4890b6e6e87c39575fd02c"
}
},
"quality": {
"equivalence": {
"status": "used",
"outcome": "non_equivalent",
"evaluator_model": "gpt-5.3-codex",
"confidence": 0.86,
"issue_type": "behavioral"
},
"code_review": {
"status": "used",
"signal": "fail",
"evaluator_model": "gpt-5.3-codex",
"confidence": 0.87,
"issue_type": "unclear",
"source_field": "code_review",
"dimension_scores": {
"correctness": {
"score_0_4": 1,
"confidence_0_1": 0.89,
"rationale": "The patch adds typename validation, but it diverges from the intended behavior: it validates at object traversal time, uses a different extension code, and emits different error semantics than requested."
},
"edge_case_handling": {
"score_0_4": 1,
"confidence_0_1": 0.82,
"rationale": "Validation is tied to presence/type of object-level `__typename` in payload, not the selected `__typename` field context, which can mis-handle cases where typename is present but not requested."
},
"introduced_bug_risk": {
"score_0_4": 1,
"confidence_0_1": 0.84,
"rationale": "Global side-channel state for object metadata and runtime object-level typename enforcement increase risk of cross-plan/state issues and unexpected nulling behavior."
},
"maintainability_idioms": {
"score_0_4": 1,
"confidence_0_1": 0.86,
"rationale": "Using a package-level `sync.Map` keyed by object pointers for schema metadata is non-idiomatic and makes reasoning about ownership/lifecycle harder than explicit struct fields."
}
},
"findings": [
{
"id": "f1",
"dimension": "correctness",
"severity": "major",
"title": "Extension code does not match intended contract",
"detail": "The patch emits `INVALID_GRAPHQL_RESPONSE` for invalid typename errors, while the intended change specifies an `INVALID_GRAPHQL` code. This is a behavioral/API mismatch for downstream error handling.",
"confidence_0_1": 0.95,
"citations": [
{
"path": "v2/pkg/engine/resolve/const.go",
"side": "new",
"start_line": 45,
"end_line": 49,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
},
{
"path": "v2/pkg/engine/resolve/resolvable.go",
"side": "new",
"start_line": 1037,
"end_line": 1045,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
}
]
},
{
"id": "f2",
"dimension": "correctness",
"severity": "major",
"title": "Validation moved to object walk instead of `__typename` field resolution",
"detail": "The implementation validates `value.__typename` during object traversal and can null objects on failure, rather than validating returned values in the `__typename` scalar selection context. This changes semantics and can trigger errors when `__typename` is not part of the selected field set.",
"confidence_0_1": 0.89,
"citations": [
{
"path": "v2/pkg/engine/resolve/resolvable.go",
"side": "new",
"start_line": 456,
"end_line": 470,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
},
{
"path": "v2/pkg/engine/resolve/resolvable.go",
"side": "new",
"start_line": 1021,
"end_line": 1045,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
}
]
},
{
"id": "f3",
"dimension": "edge_case_handling",
"severity": "major",
"title": "Error payload omits subgraph source context required by the change",
"detail": "Invalid typename errors are generated without datasource/source attribution; intended behavior requires descriptive subgraph-origin messaging for federated responses.",
"confidence_0_1": 0.82,
"citations": [
{
"path": "v2/pkg/engine/resolve/resolvable.go",
"side": "new",
"start_line": 1037,
"end_line": 1045,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
}
]
},
{
"id": "f4",
"dimension": "maintainability_idioms",
"severity": "major",
"title": "Global pointer-keyed metadata map introduces hidden lifecycle/state complexity",
"detail": "Storing possible typename metadata in a package-level `sync.Map` keyed by `*Object` creates implicit shared state and potential leaks/stale metadata concerns, instead of explicit schema metadata on the node itself.",
"confidence_0_1": 0.9,
"citations": [
{
"path": "v2/pkg/engine/resolve/node_object.go",
"side": "new",
"start_line": 8,
"end_line": 11,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
},
{
"path": "v2/pkg/engine/resolve/node_object.go",
"side": "new",
"start_line": 69,
"end_line": 90,
"anchor_kind": "diff",
"head_sha": "f45cb015e9a28689c83e886b746d21abbd030d37"
}
]
}
],
"overall": {
"overall_score_0_100": 25,
"signal": "fail",
"confidence_overall": 0.87,
"summary": "The patch partially addresses typename validation and quote escaping, but it likely does not satisfy the intended change due to mismatched error contract (`INVALID_GRAPHQL`), different validation location/semantics, and non-idiomatic global metadata handling."
}
},
"behavioral_robustness": {
"status": "used",
"probe_status": "used",
"gold_pass_candidate_pass_count": 1,
"agreement_rate": 1,
"coverage_delta_status": "unavailable",
"coverage_delta_reason": "coverage_adapter_unavailable",
"mutation_lite_status": "unavailable",
"mutation_lite_reason": "mutation_runner_unavailable",
"notes": "derived from command-level outcomes; no accepted generated probes"
},
"footprint_risk": {
"status": "used",
"reason": "none",
"version": "v1",
"level": "medium",
"score": 0.34736249848833334,
"flag": false,
"severe_flag": false
},
"cost": {
"status": "tokens_only",
"token_status": "present",
"cache_token_status": "present",
"total_input_tokens": 4553022,
"total_output_tokens": 16513,
"total_tokens": 4569535,
"uncached_input_tokens": 153918,
"cache_read_input_tokens": 4399104,
"cached_input_tokens": 4399104,
"total_cost_usd": 2.639492,
"token_source": "openai_cached_tokens_usage",
"pricing_model_key": "gpt-5.4",
"pricing_mode": "cache_aware"
}
},
"summary": {
"task": {
"matrix_status": "pass_with_warn",
"tests_outcome": "pass",
"tests_unknown_cause": null,
"lane_report_lane": "lane_unknown",
"lane_report_source": "lane_unknown",
"lane_report_reasons": [
"lane_unknown"
],
"cache_hit": false,
"cache_miss_reason": "missing_pinned_dependencies",
"setup_ms_saved": null,
"pinned_bytes": null,
"environment_group_id": "eg_46b6a6c82a4890b6e6e87c39575fd02c",
"equivalence_status": "used",
"equivalence_outcome": "non_equivalent",
"code_review_status": "used",
"code_review_signal": "fail",
"behavioral_robustness_status": "used",
"coverage_delta_status": "unavailable",
"mutation_lite_status": "unavailable",
"probe_accepted_candidates": null,
"probe_accepted_commands": null,
"probe_agreement_rate": 1,
"probe_gold_pass_candidate_pass_count": 1,
"probe_gold_pass_candidate_fail_count": null,
"probe_review_required_count": null,
"flags": [
"equiv_warn",
"review_warn"
],
"tb_is_resolved": true,
"tb_failure_mode": "unset",
"tb_total_input_tokens": 4553022,
"tb_total_output_tokens": 16513,
"tb_total_tokens": 4569535,
"tb_uncached_input_tokens": 153918,
"tb_cache_creation_input_tokens": null,
"tb_cache_read_input_tokens": 4399104,
"tb_cached_input_tokens": 4399104,
"token_status": "present",
"cache_token_status": "present",
"token_source": "openai_cached_tokens_usage",
"task_cost": 2.639492,
"cost_status": "present",
"cost_pricing_mode": "cache_aware",
"pricing_model_key": "gpt-5.4",
"equiv_rescue_policy": "on",
"rescue_candidate": false,
"rescue_eligible": false,
"rescue_decision": "not_candidate",
"publish_include_in_leaderboard": true,
"publish_exclusion_reasons": [],
"publish_weak_signal_risk": false,
"tests_only_outcome": 1,
"rescue_aware_outcome": 1,
"partial_score": 1,
"partial_score_numerator": 1,
"partial_score_denominator": 1,
"partial_score_level": "command",
"partial_score_provenance": "fallback_command_level",
"partial_score_reason": "test_case_detail_unavailable",
"partial_score_unknown_count": 0,
"footprint_risk_status": "used",
"footprint_risk_reason": "none",
"footprint_risk_level": "medium",
"footprint_risk_score": 0.34736249848833334,
"footprint_risk_flag": false,
"footprint_risk_severe_flag": false
},
"run": {
"model": "gpt-5.4",
"requested_model": "gpt-5.4",
"run_id": "2026-02-28__13-38-10__gpt-5-4",
"passRate": 0.896551724137931,
"ciLow": 0.7586206896551724,
"ciHigh": 1,
"effectiveN": 29,
"tier": 1,
"validation_counts": {
"fail_guardrail": 1,
"fail_high_conf": 2,
"fail_infra": 0,
"fail_likely_equiv": 1,
"fail_with_diag": 0,
"missing": 0,
"pass": 5,
"pass_with_warn": 20
}
},
"run_metadata": {
"agent_name": "codex",
"model_name": "gpt-5.4",
"start_time": "2026-03-06T11:06:36.354605+00:00",
"end_time": "2026-03-06T13:16:38.693205+00:00",
"commit_hash": "8f1bc9fe2c3cb8a14c254de41e9577a2ce352ac2",
"accuracy": 0.7586206896551724,
"reasoning_effort": "medium"
}
},
"evidence": {
"task_detail": {
"path": "validation/gpt-5-4/flux-pr-891/task_detail.v1.json",
"exists": true
},
"trajectory": {
"path": "validation/gpt-5-4/flux-pr-891/trajectory.v1.json",
"exists": true,
"non_empty": true,
"size_bytes": 15224
},
"validation": {
"path": "validation/gpt-5-4/flux-pr-891/validation.json",
"exists": true,
"non_empty": true,
"size_bytes": 137601
},
"results": {
"path": "runs/2026-02-28__13-38-10__gpt-5-4/results.json",
"exists": true,
"non_empty": true,
"size_bytes": 61729
},
"run_metadata": {
"path": "runs/2026-02-28__13-38-10__gpt-5-4/run_metadata.json",
"exists": true,
"non_empty": true,
"size_bytes": 1623
},
"agent_patch": {
"path": "runs/2026-02-28__13-38-10__gpt-5-4/flux-pr-891/flux-pr-891.1-of-1.2026-02-28__13-38-10__gpt-5-4/agent-logs/agent.patch",
"exists": true,
"non_empty": true,
"size_bytes": 15442
},
"summary": {
"path": "reports/summary.json",
"exists": true,
"non_empty": true,
"size_bytes": 269968
},
"manifest": {
"path": "manifest.json",
"exists": true,
"non_empty": true,
"size_bytes": 677
},
"validation_inputs": {
"required_for_rescue_ok": true,
"gold_patch": {
"exists": true,
"non_empty": true,
"size_bytes": 10055,
"sha256": "a0195599ed9d85dd37db5345cc3e3ce6bb0437b570d6cf28e6ceaeadeaf16ba4",
"required_for_rescue": true
},
"repo_tarball": {
"exists": true,
"non_empty": true,
"size_bytes": 4127754,
"sha256": "1dcc65506817b5e10093dccef87843885126d7d0962a028eb4c705e973d29f7c",
"required_for_rescue": true
},
"agent_patch": {
"path": "runs/2026-02-28__13-38-10__gpt-5-4/flux-pr-891/flux-pr-891.1-of-1.2026-02-28__13-38-10__gpt-5-4/agent-logs/agent.patch",
"exists": true,
"non_empty": true,
"size_bytes": 15442,
"sha256": "d494876a6f3f84dcd9b0c8f82b495472c7450c0bf436dd5f0f8679bc5d8890b1",
"required_for_rescue": true
}
}
},
"provenance": {
"identity": [
{
"source": "manifest",
"available": true
},
{
"source": "tb_results",
"available": true
}
],
"decision": [
{
"source": "validation",
"available": true
}
],
"tests": [
{
"source": "validation",
"available": true
},
{
"source": "tb_results",
"available": true
}
],
"quality": {
"equivalence": [
{
"source": "validation",
"available": true
}
],
"code_review": [
{
"source": "validation",
"available": true
}
],
"behavioral_robustness": [
{
"source": "validation",
"available": true
}
],
"footprint_risk": [
{
"source": "validation",
"available": true
}
],
"cost": [
{
"source": "validation",
"available": true
},
{
"source": "summary",
"available": true
}
]
},
"summary": {
"task": [
{
"source": "summary",
"available": true
}
],
"run": [
{
"source": "summary",
"available": true
}
],
"run_metadata": [
{
"source": "run_metadata",
"available": true
}
]
},
"evidence": [
{
"source": "filesystem",
"available": true
}
]
}
}