{ "schema_version": "alliance-calibration-snapshot.v1", "run_id": "g1-candidate-prompt-1.4.0-2026-08-07", "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "benchmark_schema": "vignette.outcome_alliance_benchmark.v1", "generated_at": "2026-08-07T05:01:16.444248Z", "predictions": [ { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "goal", "value": null, "confidence": null, "evidence_turn_indices": [], "status": "error", "error_code": "no_structured_output", "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "fbbc734e-d881-4c04-b42d-1db652a7e034", "attempt_count": 3, "prior_model_run_ids": [ "94ff0f1c-a262-4dce-bb8d-ec85575ae1fe", "6152ca37-a579-4b1c-80a0-06bbeba2b22f" ], "prior_error_codes": [ "no_structured_output", "no_structured_output" ], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "0349809a6e585d3c178310e94ef79e11ed967ce3409f3961517cb0fc15b1d7dd" }, { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "task", "value": null, "confidence": null, "evidence_turn_indices": [], "status": "error", "error_code": "no_structured_output", "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "fbbc734e-d881-4c04-b42d-1db652a7e034", "attempt_count": 3, "prior_model_run_ids": [ "94ff0f1c-a262-4dce-bb8d-ec85575ae1fe", "6152ca37-a579-4b1c-80a0-06bbeba2b22f" ], "prior_error_codes": [ "no_structured_output", "no_structured_output" ], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "0349809a6e585d3c178310e94ef79e11ed967ce3409f3961517cb0fc15b1d7dd" }, { "case_id": "oas-g0-001", "perspective": "independent_observer", "dimension": "bond", "value": null, "confidence": null, "evidence_turn_indices": [], "status": "error", "error_code": "no_structured_output", "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "fbbc734e-d881-4c04-b42d-1db652a7e034", "attempt_count": 3, "prior_model_run_ids": [ "94ff0f1c-a262-4dce-bb8d-ec85575ae1fe", "6152ca37-a579-4b1c-80a0-06bbeba2b22f" ], "prior_error_codes": [ "no_structured_output", "no_structured_output" ], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "0349809a6e585d3c178310e94ef79e11ed967ce3409f3961517cb0fc15b1d7dd" }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "goal", "value": 0.4, "confidence": 0.15, "evidence_turn_indices": [ 0 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "c45b6daf-9fc7-4a63-9090-450de5cf1e23", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "6754fd04565d42b83214b23a8938a55a2275f8e28bcd94f0caa027a43ca13fab" }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "task", "value": 0.15, "confidence": 0.85, "evidence_turn_indices": [ 0, 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "c45b6daf-9fc7-4a63-9090-450de5cf1e23", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "6754fd04565d42b83214b23a8938a55a2275f8e28bcd94f0caa027a43ca13fab" }, { "case_id": "oas-g0-002", "perspective": "independent_observer", "dimension": "bond", "value": 0.3, "confidence": 0.3, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "c45b6daf-9fc7-4a63-9090-450de5cf1e23", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "6754fd04565d42b83214b23a8938a55a2275f8e28bcd94f0caa027a43ca13fab" }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "goal", "value": 0.25, "confidence": 0.2, "evidence_turn_indices": [ 0 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "614df722-b37a-453e-a0e4-2fef68add0ef", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "6012f42d024bc0a8a07c5ffbc647eef6424c3e653291ebfd51d067e4b81f6362" }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "task", "value": 0.3, "confidence": 0.3, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "614df722-b37a-453e-a0e4-2fef68add0ef", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "6012f42d024bc0a8a07c5ffbc647eef6424c3e653291ebfd51d067e4b81f6362" }, { "case_id": "oas-g0-003", "perspective": "independent_observer", "dimension": "bond", "value": 0.35, "confidence": 0.3, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "614df722-b37a-453e-a0e4-2fef68add0ef", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "6012f42d024bc0a8a07c5ffbc647eef6424c3e653291ebfd51d067e4b81f6362" }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "goal", "value": 0.2, "confidence": 0.35, "evidence_turn_indices": [ 0, 1 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "0b8dc9df-7c3b-460b-8580-dd3a76c0615d", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "49a0a720dd80edee39b128619fbe9a052139ef889acde66c8e189dc5d662d736" }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "task", "value": 0.15, "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": "1.4.0", "model_run_id": "0b8dc9df-7c3b-460b-8580-dd3a76c0615d", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "49a0a720dd80edee39b128619fbe9a052139ef889acde66c8e189dc5d662d736" }, { "case_id": "oas-g0-005", "perspective": "client_agent_report", "dimension": "bond", "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": "1.4.0", "model_run_id": "0b8dc9df-7c3b-460b-8580-dd3a76c0615d", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "49a0a720dd80edee39b128619fbe9a052139ef889acde66c8e189dc5d662d736" }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "goal", "value": 0.65, "confidence": 0.55, "evidence_turn_indices": [ 0, 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "0833ec44-a429-4fae-bbfc-30de2f0fb9ca", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "36a6237c838477ebd684f7d623af80272d241d81f1151b3df703821a38a4db27" }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "task", "value": 0.78, "confidence": 0.6, "evidence_turn_indices": [ 1, 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "0833ec44-a429-4fae-bbfc-30de2f0fb9ca", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "36a6237c838477ebd684f7d623af80272d241d81f1151b3df703821a38a4db27" }, { "case_id": "oas-g0-006", "perspective": "client_agent_report", "dimension": "bond", "value": 0.6, "confidence": 0.5, "evidence_turn_indices": [ 1, 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "0833ec44-a429-4fae-bbfc-30de2f0fb9ca", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "36a6237c838477ebd684f7d623af80272d241d81f1151b3df703821a38a4db27" }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "goal", "value": 0.15, "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": "1.4.0", "model_run_id": "84f8f86a-7db6-4511-8209-02dc9d1c444d", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "236c869a1b7c94810f9a7a838ac6e2e212647105ff7f229051ff12ccf9c98b54" }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "task", "value": 0.1, "confidence": 0.55, "evidence_turn_indices": [ 1, 2 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "84f8f86a-7db6-4511-8209-02dc9d1c444d", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "236c869a1b7c94810f9a7a838ac6e2e212647105ff7f229051ff12ccf9c98b54" }, { "case_id": "oas-g0-007", "perspective": "client_agent_report", "dimension": "bond", "value": 0.08, "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": "1.4.0", "model_run_id": "84f8f86a-7db6-4511-8209-02dc9d1c444d", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "236c869a1b7c94810f9a7a838ac6e2e212647105ff7f229051ff12ccf9c98b54" }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "goal", "value": 0.15, "confidence": 0.7, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "16455942-5984-446d-939b-e4e1fa29f128", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "task", "value": 0.3, "confidence": 0.55, "evidence_turn_indices": [ 0, 1, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "16455942-5984-446d-939b-e4e1fa29f128", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "client_agent_report", "dimension": "bond", "value": 0.72, "confidence": 0.65, "evidence_turn_indices": [ 0, 1 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "16455942-5984-446d-939b-e4e1fa29f128", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "b4891122ed4fb44392954ee4981fccaae99db38ba7757418641d676853ca985b", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "goal", "value": 0.25, "confidence": 0.5, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "f21cbf10-6ce5-4cae-9ea9-f93f30bc200f", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "task", "value": 0.25, "confidence": 0.45, "evidence_turn_indices": [ 2, 3 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "f21cbf10-6ce5-4cae-9ea9-f93f30bc200f", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" }, { "case_id": "oas-g0-008", "perspective": "independent_observer", "dimension": "bond", "value": 0.65, "confidence": 0.5, "evidence_turn_indices": [ 0, 1 ], "status": "ready", "error_code": null, "provider": "claude_cli", "model": "claude-opus-4-8", "prompt_bundle_version": "1.4.0", "model_run_id": "f21cbf10-6ce5-4cae-9ea9-f93f30bc200f", "attempt_count": 1, "prior_model_run_ids": [], "prior_error_codes": [], "prompt_bundle_hash": "9e2b925c94949101b527beb5c39c46469655f81742b8758f0d36bc368029d2a4", "input_evidence_hash": "dab09723f26a5e58eb3b321abb8f4aec01d25063e6aaa198dffdcbd1ca7ad360" } ] }