{ "schema_version": "alliance-calibration-snapshot.v1", "run_id": "g1-candidate-prompt-1.2.0-readskewfix-2026-08-07", "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "benchmark_schema": "vignette.outcome_alliance_benchmark.v1", "generated_at": "2026-08-07T07:56:52.784424Z", "predictions": [ { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "goal", "value": 0.1, "confidence": 0.75, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "0b1922a3-c989-412d-89f7-42e37811afea", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "0349809a6e585d3c178310e94ef79e11ed967ce3409f3961517cb0fc15b1d7dd" }, { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "task", "value": 0.1, "confidence": 0.75, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "0b1922a3-c989-412d-89f7-42e37811afea", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "0349809a6e585d3c178310e94ef79e11ed967ce3409f3961517cb0fc15b1d7dd" }, { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "bond", "value": 0.3, "confidence": 0.25, "evidence_turn_indices": [ 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "0b1922a3-c989-412d-89f7-42e37811afea", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "0349809a6e585d3c178310e94ef79e11ed967ce3409f3961517cb0fc15b1d7dd" }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "goal", "value": 0.2, "confidence": 0.25, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "795382b9-cf8d-49e5-817d-654a562ffde9", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "6754fd04565d42b83214b23a8938a55a2275f8e28bcd94f0caa027a43ca13fab" }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "task", "value": 0.15, "confidence": 0.6, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "795382b9-cf8d-49e5-817d-654a562ffde9", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "6754fd04565d42b83214b23a8938a55a2275f8e28bcd94f0caa027a43ca13fab" }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "bond", "value": 0.2, "confidence": 0.35, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "795382b9-cf8d-49e5-817d-654a562ffde9", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "6754fd04565d42b83214b23a8938a55a2275f8e28bcd94f0caa027a43ca13fab" }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "goal", "value": 0.3, "confidence": 0.2, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "a13a14a8-f9c8-46a2-98f1-041a3e4ca923", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "6012f42d024bc0a8a07c5ffbc647eef6424c3e653291ebfd51d067e4b81f6362" }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "task", "value": 0.3, "confidence": 0.15, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "a13a14a8-f9c8-46a2-98f1-041a3e4ca923", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "6012f42d024bc0a8a07c5ffbc647eef6424c3e653291ebfd51d067e4b81f6362" }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "bond", "value": 0.4, "confidence": 0.3, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "a13a14a8-f9c8-46a2-98f1-041a3e4ca923", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "6012f42d024bc0a8a07c5ffbc647eef6424c3e653291ebfd51d067e4b81f6362" }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "goal", "value": 0.1, "confidence": 0.25, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "11b3051d-5045-429f-92ef-f5db1f6d6254", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "49a0a720dd80edee39b128619fbe9a052139ef889acde66c8e189dc5d662d736" }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "task", "value": 0.1, "confidence": 0.3, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "11b3051d-5045-429f-92ef-f5db1f6d6254", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "49a0a720dd80edee39b128619fbe9a052139ef889acde66c8e189dc5d662d736" }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "bond", "value": 0.1, "confidence": 0.65, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "11b3051d-5045-429f-92ef-f5db1f6d6254", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "49a0a720dd80edee39b128619fbe9a052139ef889acde66c8e189dc5d662d736" }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "goal", "value": 0.55, "confidence": 0.4, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "6b6110a8-0a25-4cd1-ac6b-019cffe819d0", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "36a6237c838477ebd684f7d623af80272d241d81f1151b3df703821a38a4db27" }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "task", "value": 0.8, "confidence": 0.8, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "6b6110a8-0a25-4cd1-ac6b-019cffe819d0", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "36a6237c838477ebd684f7d623af80272d241d81f1151b3df703821a38a4db27" }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "bond", "value": 0.5, "confidence": 0.4, "evidence_turn_indices": [ 0, 1, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "6b6110a8-0a25-4cd1-ac6b-019cffe819d0", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "36a6237c838477ebd684f7d623af80272d241d81f1151b3df703821a38a4db27" }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "goal", "value": 0.1, "confidence": 0.25, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "29c2c4d3-a069-48ae-ad04-93c184d41b5a", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "236c869a1b7c94810f9a7a838ac6e2e212647105ff7f229051ff12ccf9c98b54" }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "task", "value": 0.05, "confidence": 0.6, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "29c2c4d3-a069-48ae-ad04-93c184d41b5a", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "236c869a1b7c94810f9a7a838ac6e2e212647105ff7f229051ff12ccf9c98b54" }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "bond", "value": 0.1, "confidence": 0.65, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "29c2c4d3-a069-48ae-ad04-93c184d41b5a", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "236c869a1b7c94810f9a7a838ac6e2e212647105ff7f229051ff12ccf9c98b54" }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "goal", "value": 0.15, "confidence": 0.55, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "468e456b-c0c1-4239-aaa9-54aab641edac", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "task", "value": 0.3, "confidence": 0.5, "evidence_turn_indices": [ 0, 1, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "468e456b-c0c1-4239-aaa9-54aab641edac", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "bond", "value": 0.7, "confidence": 0.55, "evidence_turn_indices": [ 0, 1 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "468e456b-c0c1-4239-aaa9-54aab641edac", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "44a5b06787cff6a3a31e734bc9a0b6fa4a3ae543635927adc324c8a3196a6933", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "goal", "value": 0.2, "confidence": 0.45, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "514fd063-8783-4473-9854-1cfbafa4a694", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "task", "value": 0.3, "confidence": 0.4, "evidence_turn_indices": [ 0, 1, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "514fd063-8783-4473-9854-1cfbafa4a694", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "bond", "value": 0.6, "confidence": 0.5, "evidence_turn_indices": [ 0, 1 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.2.0", "model_run_id": "514fd063-8783-4473-9854-1cfbafa4a694", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "8e5ad200cb35ca3c01cca12c12cecf8c6237c9ad1af25ddfd9b766ad8a538fd8", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" } ] }