{ "schema_version": "alliance-calibration-comparison.v1", "generated_at": "2026-08-06 12:23:42.409719+00:00", "baseline": { "run": { "run_id": "oas-g0-human-gold-v1", "provider": "human_gold", "model": "oas-g0-directional-reference", "prompt_bundle_version": "not-applicable", "generated_at": "2026-08-06 12:23:42.389503+00:00" }, "expected_count": 9, "ready_count": 9, "missing_or_failed_count": 0, "error_count": 0, "direction_accuracy": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0, "dimension_means": { "bond": 0.35000000000000003, "goal": 0.2, "task": 0.39999999999999997 }, "rows": [ { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "goal", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "task", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "bond", "expected_direction": "drop", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "bond", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "task", "expected_direction": "rise", "status": "ready", "error_code": null, "direction_match": true, "value": 0.8, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "bond", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "bond", "expected_direction": "high", "status": "ready", "error_code": null, "direction_match": true, "value": 0.8, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "goal", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "task", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.2, "confidence": 1.0, "evidence_precision": 1.0, "evidence_recall": 1.0 } ] }, "candidate": { "run": { "run_id": "g1-candidate-2026-08-06", "provider": "claude_cli", "model": "mixed", "prompt_bundle_version": "1.0.0", "generated_at": "2026-08-06 12:21:16.722685+00:00" }, "expected_count": 9, "ready_count": 8, "missing_or_failed_count": 1, "error_count": 1, "direction_accuracy": 0.6666666666666666, "evidence_precision": 0.75, "evidence_recall": 0.8333333333333334, "dimension_means": { "bond": 0.2925, "goal": 0.1, "task": 0.325 }, "rows": [ { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "goal", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.1, "confidence": 0.7, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "task", "expected_direction": "low", "status": "error", "error_code": "agent_validationerror", "direction_match": false, "value": null, "confidence": null, "evidence_precision": 0.0, "evidence_recall": 0.0 }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "bond", "expected_direction": "drop", "status": "ready", "error_code": null, "direction_match": true, "value": 0.4, "confidence": 0.3, "evidence_precision": 0.6666666666666666, "evidence_recall": 1.0 }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "bond", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.12, "confidence": 0.75, "evidence_precision": 0.3333333333333333, "evidence_recall": 1.0 }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "task", "expected_direction": "rise", "status": "ready", "error_code": null, "direction_match": false, "value": 0.55, "confidence": 0.5, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "bond", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.1, "confidence": 0.85, "evidence_precision": 0.3333333333333333, "evidence_recall": 1.0 }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "bond", "expected_direction": "high", "status": "ready", "error_code": null, "direction_match": false, "value": 0.55, "confidence": 0.35, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "goal", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.1, "confidence": 0.45, "evidence_precision": 1.0, "evidence_recall": 1.0 }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "task", "expected_direction": "low", "status": "ready", "error_code": null, "direction_match": true, "value": 0.1, "confidence": 0.3, "evidence_precision": 1.0, "evidence_recall": 1.0 } ] }, "drift": { "comparable_predictions": 8, "mean_absolute_score_delta": 0.14750000000000002, "dimension_mean_delta": { "bond": -0.0575, "goal": -0.1, "task": -0.175 }, "direction_accuracy_delta": -0.33333333333333337, "evidence_recall_delta": -0.16666666666666663 } }