{ "ofone_version": "0.6.0", "mode": "Audit", "charter": { "objective": "Audit a public-sector AI-assisted decision process before release.", "scope": [ "policy design", "model performance evidence", "rights and legitimacy", "human review gates", "operational monitoring" ], "horizon": "one policy cycle before live release", "stakes": [ "rights", "public-policy legitimacy", "operational safety", "legal exposure", "public trust" ], "risk_tier": "high", "movement_jobs": [ "BOUND", "WARN", "GATE" ] }, "adapter_projection": { "primary": "hybrid", "axes": { "strategic-agentic": [ "institutional incentives", "caseworker behavior", "review authority" ], "scientific-explanatory": [ "model performance", "subgroup impact", "measurement validity" ], "formal": [ "policy rules", "eligibility logic", "audit constraints" ], "normative-evaluative": [ "rights", "legitimacy", "contestability", "distributional harm" ] }, "movement_jobs": [ "BOUND", "LINK", "EVALUATE" ] }, "scene": { "scene_id": "S1", "frames": [ { "frame_id": "F1", "type": "normative", "assumptions": [ "affected people deserve notice, explanation, contestability, and meaningful review" ] }, { "frame_id": "F2", "type": "causal", "assumptions": [ "AI assistance can alter caseworker behavior and error propagation" ] }, { "frame_id": "F3", "type": "evidential", "assumptions": [ "model evidence must match the deployment population and decision workflow" ] }, { "frame_id": "F4", "type": "strategic", "assumptions": [ "public agencies need accountable release authority and visible legitimacy" ] } ], "tokens": [ { "token_id": "K1", "kind": "entity", "label": "public-sector organization" }, { "token_id": "K2", "kind": "entity", "label": "affected people" }, { "token_id": "K3", "kind": "variable", "label": "deployment-population model performance" }, { "token_id": "K4", "kind": "constraint", "label": "rights, legitimacy, and contestability requirements" }, { "token_id": "K5", "kind": "variable", "label": "human review and override behavior" }, { "token_id": "K6", "kind": "uncertainty", "label": "subgroup impact and appeal accessibility" } ], "state_variables": [ "decision outcome", "model error rate", "subgroup error rate", "appeal outcome", "caseworker override rate", "review gate status" ], "observed_variables": [ "public-sector AI-assisted decision proposal", "performance, rights, legitimacy, review, and operational constraints", "benchmark requirement for separated map objects" ], "hidden_variables": [ "automation bias", "unequal data quality", "appeal accessibility", "symbolic review risk", "stakeholder trust" ], "movement_jobs": [ "BOUND", "WARN", "GATE" ], "subscenes": [ { "subscene_id": "SS1", "purpose": "evidence_acquisition", "frames": [ "F3" ], "tokens": [ "K3", "K6" ], "entry_conditions": [ "model evidence is absent or not deployment-population specific" ], "exit_conditions": [ "model performance, subgroup impact, and appeal-access evidence are recorded" ], "parent_scene": "S1", "movement_jobs": [ "BOUND", "LINK" ] }, { "subscene_id": "SS2", "purpose": "stakeholder_context", "frames": [ "F1" ], "tokens": [ "K2", "K4", "K6" ], "entry_conditions": [ "public decision affects rights or access to public services" ], "exit_conditions": [ "notice, explanation, contestability, and affected-party exposure are reviewed" ], "parent_scene": "S1", "movement_jobs": [ "BOUND", "WARN" ] }, { "subscene_id": "SS3", "purpose": "review_gate", "frames": [ "F2", "F4" ], "tokens": [ "K1", "K5" ], "entry_conditions": [ "agency considers moving from design to live release" ], "exit_conditions": [ "named reviewers approve, block, or return the policy for evidence" ], "parent_scene": "S1", "movement_jobs": [ "BOUND", "GATE" ] } ] }, "evidence": [ { "evidence_id": "E1", "source": "file", "span_or_locator": "benchmarks/cases/public-sector-ai-policy-audit.md", "provenance": "frozen benchmark case file", "recency": "current", "reliability": "medium", "permission": "public", "content_hash": "sha256:eafc104563df8651821c28617ce84dc0c43119bd2c09d3fedd2e5fcba52f3178", "retrieved_at": "2026-05-18T04:34:00Z", "extract": "A public-sector organization is considering an AI-assisted decision process with performance, rights, legitimacy, review, and operational constraints.", "source_owner": "OfOne benchmark suite", "chain_of_custody": "Local benchmark run read the frozen case file from the repository before artifact construction.", "supports": [ "C1", "C2", "C3", "C4" ], "risks": [ "scenario_level_evidence", "no_external_policy_dossier" ], "movement_jobs": [ "GROUND", "WARN" ] }, { "evidence_id": "E2", "source": "file", "span_or_locator": "benchmarks/runs/2026-05-17-batch-01/prompts/full_ofone.md", "provenance": "frozen full-OfOne benchmark prompt", "recency": "current", "reliability": "medium", "permission": "public", "content_hash": "sha256:613afac8909b34accb57fd2c24217bb59a28a45860e32f36f6ec5f7f4ab5587e", "retrieved_at": "2026-05-18T04:34:00Z", "extract": "Full-OfOne arm must produce artifact JSON, validator result, rendering, and patch report while preserving case fidelity.", "source_owner": "OfOne benchmark suite", "chain_of_custody": "Local benchmark run computed the full-OfOne prompt hash and included it in benchmark_trace.", "supports": [ "C4" ], "risks": [ "benchmark_prompt_not_domain_source" ], "movement_jobs": [ "GROUND", "LINK" ] }, { "evidence_id": "E3", "source": "observation", "span_or_locator": "local benchmark case interpretation", "provenance": "local Codex benchmark execution", "recency": "current", "reliability": "medium", "permission": "internal", "content_hash": "sha256:f8c670fa5a6bf20575d0c23f00b5485f56718e24cf53acec929ea1ebe8fcb079", "retrieved_at": "2026-05-18T04:34:00Z", "extract": "The case requires separation of model evidence, policy claims, stakeholder exposure, gates, and review logs.", "source_owner": "local benchmark executor", "chain_of_custody": "Executor translated the frozen benchmark prompt into typed OfOne objects without using other benchmark arm outputs.", "supports": [ "C2", "C3", "C4" ], "risks": [ "unblinded_local_execution" ], "movement_jobs": [ "GROUND", "GATE" ] } ], "claims": [ { "claim_id": "C1", "text": "The proposal is a high-stakes public-sector AI-assisted decision process, not a low-risk automation.", "type": "descriptive", "supports": [ "E1" ], "contradicts": [], "depends_on": [ "K1", "K2" ], "confidence": { "level": "high", "basis": [ "provenance", "adapter_fit" ], "failure_modes": [ "case_abstraction", "specific_program_missing" ] }, "status": "active", "review_gate": true, "movement_jobs": [ "CLAIM", "BOUND", "WARN" ] }, { "claim_id": "C2", "text": "Rights and legitimacy exposure require human review gates before live release.", "type": "normative", "supports": [ "E1", "E3" ], "contradicts": [], "depends_on": [ "K2", "K4" ], "confidence": { "level": "high", "basis": [ "provenance", "adapter_fit", "hidden_variable_risk" ], "failure_modes": [ "jurisdiction_specific_rule_missing", "symbolic_review" ] }, "status": "active", "review_gate": true, "movement_jobs": [ "CLAIM", "GATE", "WARN" ] }, { "claim_id": "C3", "text": "Model performance, subgroup impact, appeal access, and override behavior are unresolved evidence dependencies.", "type": "operational", "supports": [ "E1", "E3" ], "contradicts": [], "depends_on": [ "K3", "K5", "K6" ], "confidence": { "level": "medium", "basis": [ "provenance", "hidden_variable_risk", "adapter_fit" ], "failure_modes": [ "missing_measurement", "selection_bias", "automation_bias" ] }, "status": "active", "review_gate": true, "movement_jobs": [ "CLAIM", "WARN", "TEST" ] }, { "claim_id": "C4", "text": "A valid audit must keep model evidence, policy claims, stakeholder exposure, gates, review logs, and final rendering separate.", "type": "operational", "supports": [ "E1", "E2", "E3" ], "contradicts": [], "depends_on": [ "K1", "K2", "K4" ], "confidence": { "level": "high", "basis": [ "provenance", "mechanism_fit", "adapter_fit" ], "failure_modes": [ "object_collapse", "review_log_omission" ] }, "status": "active", "review_gate": true, "movement_jobs": [ "CLAIM", "LINK", "EVALUATE" ] } ], "unknowns": [ { "unknown_id": "U1", "kind": "missing_measurement", "description": "No deployment-population performance or subgroup-impact evidence is present.", "blocks": [ "O1", "O2", "R1" ], "resolution_move": "Run model validation and subgroup-impact audit on the intended deployment population.", "status": "open", "movement_jobs": [ "WARN", "GATE", "TEST" ] }, { "unknown_id": "U2", "kind": "missing_evidence", "description": "No notice, explanation, appeal-access, or affected-party consultation evidence is present.", "blocks": [ "O1", "R1" ], "resolution_move": "Document contestability, appeal access, notice, explanation, and stakeholder exposure review.", "status": "open", "movement_jobs": [ "WARN", "GATE", "TEST" ] }, { "unknown_id": "U3", "kind": "missing_evidence", "description": "No completed review-log decision or accountable release authority is present.", "blocks": [ "O1", "O2", "R1" ], "resolution_move": "Assign reviewers and record gate decisions before any release.", "status": "open", "movement_jobs": [ "WARN", "GATE", "TEST" ] } ], "kill_tests": [ { "test_id": "KT1", "target": "C2", "test_type": "stakeholder_objection", "condition": "Gate escalation or stakeholder objection shows the proposed review design lacks meaningful contestability.", "falsifies": [ "C2" ], "movement_jobs": [ "TEST", "WARN", "GATE" ] }, { "test_id": "KT2", "target": "C3", "test_type": "measurement", "condition": "Source contradiction or subgroup measurement failure shows model evidence does not match the deployment population.", "falsifies": [ "C3" ], "movement_jobs": [ "TEST", "WARN" ] }, { "test_id": "KT3", "target": "C4", "test_type": "adapter_conflict", "condition": "Adapter conflict collapses evidence, policy claims, stakeholder exposure, gates, or review logs into one undifferentiated recommendation.", "falsifies": [ "C4" ], "movement_jobs": [ "TEST", "EVALUATE" ] } ], "criteria": [ { "criterion_id": "CR1", "name": "Rights and public-policy gate", "kind": "threshold", "priority": "must", "threshold": "No live release before rights, public-policy, and contestability gates are reviewed.", "owned_by": [ "A1", "A3" ], "movement_jobs": [ "EVALUATE", "GATE" ] }, { "criterion_id": "CR2", "name": "Model evidence fit", "kind": "threshold", "priority": "must", "threshold": "Model evidence must cover the deployment population, subgroup impact, and operational workflow.", "owned_by": [ "A2" ], "movement_jobs": [ "EVALUATE", "TEST" ] }, { "criterion_id": "CR3", "name": "Review-log accountability", "kind": "constraint", "priority": "must", "threshold": "Gate decisions must name reviewer authority and produce review-log evidence before release.", "owned_by": [ "A1", "A2" ], "movement_jobs": [ "GATE", "WARN" ] }, { "criterion_id": "CR4", "name": "Legitimacy and affected-party exposure", "kind": "objective", "priority": "should", "threshold": "Prefer moves that surface affected-party exposure and preserve appeal access.", "owned_by": [ "A3" ], "movement_jobs": [ "EVALUATE", "WARN" ] } ], "tradeoff_surface": { "surface_id": "TS1", "options": [ "O1", "O2" ], "criteria": [ "CR1", "CR2", "CR3", "CR4" ], "dominant_option": "O1", "why": [ "CR1", "CR2", "CR3" ], "reversal_conditions": [ "U1", "U2", "U3", "T1", "T2" ], "movement_jobs": [ "EVALUATE", "TRIGGER" ] }, "actors": [ { "actor_id": "A1", "label": "public agency decision owner", "role": "decision_owner", "incentives": [ "deliver public service while avoiding invalid release" ], "exposures": [ "legal exposure", "public trust", "rights" ], "authority": "approve", "legitimacy_basis": "assigned accountability for public-sector release decision", "movement_jobs": [ "BOUND", "GATE", "WARN" ] }, { "actor_id": "A2", "label": "technical and policy reviewer", "role": "reviewer", "incentives": [ "verify model evidence and operational controls" ], "exposures": [ "operational safety", "model validity", "review quality" ], "authority": "block", "legitimacy_basis": "assigned technical and policy review responsibility", "movement_jobs": [ "BOUND", "TEST", "GATE" ] }, { "actor_id": "A3", "label": "affected-party representative", "role": "affected_party", "incentives": [ "avoid erroneous or inaccessible public decision outcomes" ], "exposures": [ "rights", "appeal access", "distributional harm" ], "authority": "observe", "legitimacy_basis": "direct exposure to the decision outcome and legitimacy claim", "movement_jobs": [ "BOUND", "WARN" ] } ], "temporal_model": { "time_horizon": "one policy cycle before live release", "decision_deadline": "before any public deployment decision", "evidence_validity_windows": [ { "evidence_id": "E1", "valid_until": "unknown", "staleness_trigger": "case dossier or policy scope changes" }, { "evidence_id": "E2", "valid_until": "unknown", "staleness_trigger": "benchmark full-OfOne prompt changes" }, { "evidence_id": "E3", "valid_until": "unknown", "staleness_trigger": "local benchmark interpretation is superseded by adjudication or external review" } ], "update_cadence": "before release, on new model evidence, on stakeholder objection, and after any review decision", "movement_jobs": [ "BOUND", "TRIGGER", "WARN" ] }, "information_value": [ { "unknown_id": "U1", "decision_impact": "high", "resolution_cost": "medium", "time_to_resolve": "2 to 6 weeks", "risk_reduction": "high", "recommended_next_query": "Collect deployment-population performance, subgroup error, calibration, and monitoring evidence.", "movement_jobs": [ "TEST", "MOVE", "EVALUATE" ] }, { "unknown_id": "U2", "decision_impact": "high", "resolution_cost": "medium", "time_to_resolve": "2 to 4 weeks", "risk_reduction": "high", "recommended_next_query": "Test notice, explanation, appeal access, and affected-party consultation evidence.", "movement_jobs": [ "TEST", "MOVE", "EVALUATE" ] }, { "unknown_id": "U3", "decision_impact": "high", "resolution_cost": "low", "time_to_resolve": "days to 2 weeks", "risk_reduction": "medium", "recommended_next_query": "Assign reviewer authority and record returned-for-evidence or blocked gate decisions.", "movement_jobs": [ "TEST", "MOVE", "GATE" ] } ], "lenses": [ { "lens_id": "LENS1", "name": "strategic agency lens", "adapter_axis": "strategic-agentic", "questions": [ "Who can approve, block, or game the release decision?" ], "claims_examined": [ "C1", "C4" ], "blind_spots": [ "agency incentives and caseworker override behavior are not evidenced" ], "contention": [ "U3 blocks release authority and review-log proof" ], "movement_jobs": [ "WARN", "TEST", "EVALUATE" ] }, { "lens_id": "LENS2", "name": "model evidence lens", "adapter_axis": "scientific-explanatory", "questions": [ "Does model evidence fit the deployment population and subgroup risks?" ], "claims_examined": [ "C3" ], "blind_spots": [ "U1 blocks model validity and subgroup impact assessment" ], "contention": [ "performance evidence is not enough unless subgroup and appeal risks are visible" ], "movement_jobs": [ "WARN", "TEST", "EVALUATE" ] }, { "lens_id": "LENS3", "name": "normative legitimacy lens", "adapter_axis": "normative-evaluative", "questions": [ "Are rights, contestability, and affected-party exposure visible before release?" ], "claims_examined": [ "C2" ], "blind_spots": [ "U2 blocks rights and legitimacy review" ], "contention": [ "public-policy release without contestability evidence is gated" ], "movement_jobs": [ "WARN", "TEST", "GATE" ] }, { "lens_id": "LENS4", "name": "formal policy-rule lens", "adapter_axis": "formal", "questions": [ "Are policy rules, audit constraints, and review-log states distinguishable?" ], "claims_examined": [ "C4" ], "blind_spots": [ "review log and gate state are not yet release evidence" ], "contention": [ "review records must not be collapsed into the recommendation text" ], "movement_jobs": [ "WARN", "EVALUATE" ] } ], "council_result": { "coverage": [ "strategic-agentic", "scientific-explanatory", "formal", "normative-evaluative" ], "missing_lenses": [ "program-specific legal counsel and affected-community validation remain outside this benchmark dossier" ], "major_dissent": [ "U1, U2, and U3 block release and rendering confidence" ], "decision_effect": "Repeat-2 council result again blocks live deployment and returns the policy for model evidence, rights/appeal evidence, stakeholder exposure review, and review-log proof.", "movement_jobs": [ "EVALUATE", "WARN", "GATE" ] }, "edges": [ { "edge_id": "X1", "from": "E1", "to": "C1", "relation_family": "evidential", "relation": "supports", "evidence_refs": [ "E1" ], "confidence": "high", "movement_jobs": [ "GROUND", "LINK" ] }, { "edge_id": "X2", "from": "E3", "to": "C4", "relation_family": "evidential", "relation": "supports", "evidence_refs": [ "E3" ], "confidence": "medium", "movement_jobs": [ "GROUND", "LINK" ] }, { "edge_id": "X3", "from": "C3", "to": "C2", "relation_family": "argumentative", "relation": "supports", "evidence_refs": [ "E1", "E3" ], "confidence": "medium", "movement_jobs": [ "LINK", "WARN" ] }, { "edge_id": "X4", "from": "K4", "to": "O1", "relation_family": "causal", "relation": "constrains", "evidence_refs": [ "E1" ], "confidence": "high", "movement_jobs": [ "LINK", "GATE" ] }, { "edge_id": "X5", "from": "C4", "to": "TS1", "relation_family": "argumentative", "relation": "supports", "evidence_refs": [ "E2", "E3" ], "confidence": "high", "movement_jobs": [ "LINK", "EVALUATE" ] }, { "edge_id": "X6", "from": "G1", "to": "O2", "relation_family": "workflow_state", "relation": "blocks", "evidence_refs": [ "E3" ], "confidence": "high", "movement_jobs": [ "GATE", "WARN" ] }, { "edge_id": "X7", "from": "CR2", "to": "TS1", "relation_family": "argumentative", "relation": "supports", "evidence_refs": [ "E1" ], "confidence": "high", "movement_jobs": [ "EVALUATE", "LINK" ] } ], "loops": [ { "loop_id": "L1", "type": "measurement", "edges": [ "X1", "X3" ], "polarity": "mixed", "delay": "medium", "gain": "unknown", "control_points": [ "deployment-population validation", "subgroup audit" ], "observable_cues": [ "performance evidence may not match affected-party population" ], "failure_mode": "aggregate model performance hides subgroup failure or appeal-access harm", "movement_jobs": [ "WARN", "LINK" ] }, { "loop_id": "L2", "type": "review", "edges": [ "X4", "X6" ], "polarity": "balancing", "delay": "short", "gain": "medium", "control_points": [ "rights gate", "review log", "release authority" ], "observable_cues": [ "gate status determines whether release remains blocked" ], "failure_mode": "review becomes symbolic and fails to block invalid deployment", "movement_jobs": [ "GATE", "WARN" ] }, { "loop_id": "L3", "type": "incentive", "edges": [ "X3", "X5" ], "polarity": "reinforcing", "delay": "medium", "gain": "medium", "control_points": [ "caseworker override monitoring", "reviewer independence" ], "observable_cues": [ "operational pressure may increase deference to model output" ], "failure_mode": "institutional pressure rewards throughput over contestability", "movement_jobs": [ "WARN", "EVALUATE" ] } ], "option_moves": [ { "option_id": "O1", "move_type": "defer", "preconditions": [ "C2", "C3", "C4" ], "expected_effects": [ "X4", "X5" ], "tradeoffs": [ "delays release", "preserves rights and legitimacy", "requires additional evidence and review work" ], "blocking_unknowns": [ "U1", "U2", "U3" ], "review_gate": "G1", "movement_jobs": [ "MOVE", "EVALUATE", "GATE" ] }, { "option_id": "O2", "move_type": "test", "preconditions": [ "C1", "C3", "C4" ], "expected_effects": [ "X1", "X2", "X7" ], "tradeoffs": [ "keeps evaluation reversible", "does not authorize live decisioning", "requires technical and policy reviewer time" ], "blocking_unknowns": [ "U1", "U3" ], "review_gate": "G2", "movement_jobs": [ "MOVE", "TEST", "GATE" ] } ], "triggers": [ { "trigger_id": "T1", "condition": "review_required", "affected_objects": [ "G1", "C2" ], "transition": "human_review", "movement_jobs": [ "TRIGGER", "GATE" ] }, { "trigger_id": "T2", "condition": "new_evidence", "affected_objects": [ "E1", "E3" ], "transition": "scoped_rerun", "movement_jobs": [ "TRIGGER", "TEST" ] }, { "trigger_id": "T3", "condition": "claim_conflict", "affected_objects": [ "C3" ], "transition": "patch", "movement_jobs": [ "TRIGGER", "WARN" ] } ], "gates": [ { "gate_id": "G1", "condition": "rights and public-policy release gate for high-risk public-sector AI deployment", "reviewer": "A1 public agency decision owner with A2 technical reviewer and A3 affected-party representation", "required_decision": "approve, block, narrow, or return for model evidence, rights evidence, and review-log proof", "status": "open", "movement_jobs": [ "GATE", "WARN" ] }, { "gate_id": "G2", "condition": "model performance and operational safety gate before controlled pilot", "reviewer": "A2 technical and policy reviewer", "required_decision": "approve test-only evaluation, block release, or require additional measurement evidence", "status": "open", "movement_jobs": [ "GATE", "TEST" ] } ], "confidence_model": { "overall": "medium", "provenance_strength": "medium", "source_independence": "low", "recency": "high", "mechanism_fit": "medium", "contradiction_load": "low", "hidden_variable_risk": "high", "adversarial_risk": "medium", "adapter_fit": "high", "movement_jobs": [ "EVALUATE", "WARN" ] }, "decision_rendering": { "rendering_id": "R1", "summary": "Repeat-2 Audit rendering: do not release the AI-assisted public-sector decision process; return it for model evidence, rights and legitimacy review, named human gates, and recorded review-log decisions.", "recommendation": "Block live deployment and continue only with a gated audit / controlled evaluation path.", "confidence": "medium", "depends_on": [ "C1", "C2", "C3", "C4", "U1", "U2", "U3", "O1", "O2", "G1", "G2", "TS1", "COUNCIL" ], "movement_jobs": [ "EVALUATE", "GATE", "MOVE" ] }, "artifact_identity": { "artifact_id": "OFONE-2026-05-18-B01-POLICY-FULL-R2", "case_id": "case-public-sector-ai-policy-audit-001", "objective_head": "public-sector AI policy audit repeat 2 with rights, legitimacy, review, and operational gates", "scope_hash": "sha256:eafc104563df8651821c28617ce84dc0c43119bd2c09d3fedd2e5fcba52f3178", "config_hash": "sha256:613afac8909b34accb57fd2c24217bb59a28a45860e32f36f6ec5f7f4ab5587e", "active_evidence_hashes": [ "sha256:eafc104563df8651821c28617ce84dc0c43119bd2c09d3fedd2e5fcba52f3178", "sha256:613afac8909b34accb57fd2c24217bb59a28a45860e32f36f6ec5f7f4ab5587e", "sha256:f8c670fa5a6bf20575d0c23f00b5485f56718e24cf53acec929ea1ebe8fcb079" ], "created_at": "2026-05-18T09:20:00Z", "status": "review_pending", "movement_jobs": [ "BOUND", "TRIGGER" ] }, "review_log": [ { "review_id": "RL1", "gate_id": "G1", "actor_id": "A1", "decision": "returned_for_evidence", "timestamp": "2026-05-18T09:20:00Z", "notes": "Repeat-2 release gate remains open because model evidence, rights evidence, appeal access, and review-log proof are incomplete.", "movement_jobs": [ "GATE", "WARN" ] }, { "review_id": "RL2", "gate_id": "G2", "actor_id": "A2", "decision": "returned_for_evidence", "timestamp": "2026-05-18T09:20:00Z", "notes": "Repeat-2 controlled evaluation still requires deployment-population performance and subgroup-impact evidence before approval.", "movement_jobs": [ "GATE", "TEST" ] } ], "benchmark_trace": { "trace_id": "BT-2026-05-17-B01-POLICY-FULL-R2", "suite_id": "ofone-v0.5-three-arm-evaluation", "cases_run": 1, "arms_run": [ "full_ofone" ], "model_families": 1, "superiority_ready": false, "diagnostics": [ "repeat-2 benchmark run only", "no empirical superiority claim supported", "Batch 01 remains below release minimums" ], "case_id": "case-public-sector-ai-policy-audit-001", "run_id": "2026-05-17-batch-01__case-public-sector-ai-policy-audit-001__full_ofone__agentic_coding__r2", "case_file": "benchmarks/cases/public-sector-ai-policy-audit.md", "case_file_sha256": "sha256:eafc104563df8651821c28617ce84dc0c43119bd2c09d3fedd2e5fcba52f3178", "prompt_file": "benchmarks/runs/2026-05-17-batch-01/prompts/full_ofone.md", "prompt_file_sha256": "sha256:613afac8909b34accb57fd2c24217bb59a28a45860e32f36f6ec5f7f4ab5587e", "input_bundle_sha256": "sha256:171b27fd434b9cf2dc349fb0c2906778e83caf37a483741f5f39e1d067fc39f4", "movement_jobs": [ "BOUND", "WARN", "TRIGGER" ] }, "validator_result": { "passed": true, "diagnostics": [ { "code": "OFONE_JSON_SCHEMA", "severity": "info", "check": "json_schema", "object_id": null, "object_type": null, "path": null, "message": "Audit artifact matches executable JSON Schema profile", "repair_hint": null }, { "code": "OFONE_ADAPTER_CONTRACT", "severity": "info", "check": "adapter_contract", "object_id": null, "object_type": null, "path": null, "message": "hybrid adapter contract loaded", "repair_hint": null }, { "code": "OFONE_ADAPTER_GATE_COVERAGE", "severity": "info", "check": "adapter_gate_coverage", "object_id": null, "object_type": null, "path": null, "message": "gate coverage present for legal, safety, public-policy, rights", "repair_hint": null }, { "code": "OFONE_BENCHMARK_TRACE", "severity": "warning", "check": "benchmark_trace", "object_id": null, "object_type": null, "path": null, "message": "benchmark_trace BT-2026-05-17-B01-POLICY-FULL-R2 is not superiority-ready", "repair_hint": null }, { "code": "OFONE_DEPENDENCY_CLOSURE", "severity": "info", "check": "dependency_closure", "object_id": null, "object_type": null, "path": null, "message": "trigger T1: C2, COUNCIL, KT1, L1, L2, L3, LENS3, O1, O2, R1, RL1, T1, TS1, X3, X4, X5, X6, X7 [includes rendering]", "repair_hint": null }, { "code": "OFONE_DEPENDENCY_CLOSURE", "severity": "info", "check": "dependency_closure", "object_id": null, "object_type": null, "path": null, "message": "trigger T2: C1, C2, C3, C4, COUNCIL, KT1, KT2, KT3, L1, L2, L3, LENS1, LENS2, LENS3, LENS4, O1, O2, OFONE-2026-05-18-B01-POLICY-FULL-R2, R1, T1, T2, T3, TEMPORAL, TS1, X1, X2, X3, X4, X5, X6, X7 [includes rendering]", "repair_hint": null }, { "code": "OFONE_DEPENDENCY_CLOSURE", "severity": "info", "check": "dependency_closure", "object_id": null, "object_type": null, "path": null, "message": "trigger T3: C3, COUNCIL, KT2, L1, L2, L3, LENS2, O1, O2, R1, T3, TS1, X3, X4, X5, X6, X7 [includes rendering]", "repair_hint": null }, { "code": "OFONE_SEMANTIC_VALIDATION", "severity": "info", "check": "semantic_validation", "object_id": null, "object_type": null, "path": null, "message": "semantic graph checks completed", "repair_hint": null } ], "checks": [ { "check": "json_schema", "passed": true, "notes": "Audit artifact matches executable JSON Schema profile", "code": "OFONE_JSON_SCHEMA", "severity": "info" }, { "check": "adapter_contract", "passed": true, "notes": "hybrid adapter contract loaded", "code": "OFONE_ADAPTER_CONTRACT", "severity": "info" }, { "check": "adapter_gate_coverage", "passed": true, "notes": "gate coverage present for legal, safety, public-policy, rights", "code": "OFONE_ADAPTER_GATE_COVERAGE", "severity": "info" }, { "check": "benchmark_trace", "passed": true, "notes": "benchmark_trace BT-2026-05-17-B01-POLICY-FULL-R2 is not superiority-ready", "code": "OFONE_BENCHMARK_TRACE", "severity": "warning" }, { "check": "dependency_closure", "passed": true, "notes": "trigger T1: C2, COUNCIL, KT1, L1, L2, L3, LENS3, O1, O2, R1, RL1, T1, TS1, X3, X4, X5, X6, X7 [includes rendering]", "code": "OFONE_DEPENDENCY_CLOSURE", "severity": "info" }, { "check": "dependency_closure", "passed": true, "notes": "trigger T2: C1, C2, C3, C4, COUNCIL, KT1, KT2, KT3, L1, L2, L3, LENS1, LENS2, LENS3, LENS4, O1, O2, OFONE-2026-05-18-B01-POLICY-FULL-R2, R1, T1, T2, T3, TEMPORAL, TS1, X1, X2, X3, X4, X5, X6, X7 [includes rendering]", "code": "OFONE_DEPENDENCY_CLOSURE", "severity": "info" }, { "check": "dependency_closure", "passed": true, "notes": "trigger T3: C3, COUNCIL, KT2, L1, L2, L3, LENS2, O1, O2, R1, T3, TS1, X3, X4, X5, X6, X7 [includes rendering]", "code": "OFONE_DEPENDENCY_CLOSURE", "severity": "info" }, { "check": "semantic_validation", "passed": true, "notes": "semantic graph checks completed", "code": "OFONE_SEMANTIC_VALIDATION", "severity": "info" } ] } }