{
 "artifact_dir": "artifacts/e81",
 "checks": [
  {
   "check": "model_call_request_ids_unique",
   "detail": 1475,
   "status": "PASS"
  },
  {
   "check": "model_call_output_hashes",
   "detail": "sha256(output_text) equals the recorded digest on every row",
   "status": "PASS"
  },
  {
   "check": "no_remote_or_mock_rows_counted_as_real",
   "detail": {
    "LOCAL_MODEL_INFERENCE": 319,
    "REAL_LOCAL_MODEL": 1156
   },
   "status": "PASS"
  },
  {
   "check": "all_calls_self_hosted",
   "detail": [
    "self_hosted_hf"
   ],
   "status": "PASS"
  },
  {
   "check": "gate_audit_row_hashes",
   "detail": "2094 rows re-hashed",
   "status": "PASS"
  },
  {
   "check": "gate_audit_every_prev_exists",
   "detail": "no row links to an unknown hash (no orphan, no spliced row)",
   "status": "PASS"
  },
  {
   "check": "every_model_call_traces_to_an_ALLOW_decision",
   "detail": "1475 calls, 0 without an ALLOW grant",
   "status": "PASS"
  },
  {
   "check": "no_EXECUTE_without_ALLOW",
   "detail": 0,
   "status": "PASS"
  },
  {
   "check": "benchmark_shape",
   "detail": "15 arms x 24 tasks = 360 rows",
   "status": "PASS"
  },
  {
   "check": "benchmark_accuracy_matches_rows",
   "detail": "every recorded arm accuracy equals the recount",
   "status": "PASS"
  },
  {
   "check": "benchmark_verdict_follows_from_rows",
   "detail": "best collaborative delta 0.0; superior=False",
   "status": "PASS"
  },
  {
   "check": "endurance_wall_clock",
   "detail": "completed-started=3600.0 s, target 3600.0 s",
   "status": "PASS"
  },
  {
   "check": "endurance_calls_span_the_hour",
   "detail": "114 calls over 3562 s, longest gap 61 s",
   "status": "PASS"
  },
  {
   "check": "security_lab_recount",
   "detail": "60/60 blocked, 0 bypassed (same-project lab)",
   "status": "PASS"
  },
  {
   "check": "capacity_stages_clean",
   "detail": [
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 9.759,
     "tasks_done": 40,
     "workers": 10
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 131.48,
     "tasks_done": 200,
     "workers": 50
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 434.674,
     "tasks_done": 400,
     "workers": 100
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 2146.473,
     "tasks_done": 1000,
     "workers": 250
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 6082.557,
     "tasks_done": 2000,
     "workers": 500
    }
   ],
   "status": "PASS"
  },
  {
   "check": "durable_stores_intact_after_host_power_loss",
   "detail": [
    {
     "bytes": 176128,
     "duplicate_effect_keys": 0,
     "integrity_check": "ok",
     "rows": {
      "checkpoints": 114,
      "dead_letters": 0,
      "effects": 116,
      "events": 470,
      "meta": 1,
      "missions": 1,
      "sqlite_sequence": 1,
      "tasks": 116
     },
     "store": "endurance/endurance.db",
     "task_states": {
      "DONE": 116
     },
     "wal_bytes": 0
    },
    {
     "bytes": 65536,
     "duplicate_effect_keys": 0,
     "integrity_check": "ok",
     "rows": {
      "checkpoints": 1,
      "dead_letters": 0,
      "effects": 6,
      "events": 25,
      "meta": 1,
      "missions": 1,
      "sqlite_sequence": 1,
      "tasks": 6
     },
     "store": "work/mission/mission.db",
     "task_states": {
      "DONE": 6
     },
     "wal_bytes": null
    }
   ],
   "status": "PASS"
  },
  {
   "check": "endurance_store_matches_report",
   "detail": {
    "db_checkpoints": 114,
    "db_effects": 116,
    "reported": 116
   },
   "status": "PASS"
  }
 ],
 "checks_not_evaluable": [],
 "checks_passed": 17,
 "checks_total": 17,
 "findings": [
  {
   "detail": "cost_report/docs say 1156 real local calls; the main ledger holds 1475 (319 early rows carry the older label LOCAL_MODEL_INFERENCE) and the endurance ledger a further 114. Total 1589.",
   "effect": "An undercount, not an overclaim. Token totals and inference seconds in the cost report already cover every ledger row, so the cost estimate is unaffected; only the call count is wrong.",
   "id": "F-01",
   "severity": "LOW",
   "title": "Reported call total undercounts the ledger"
  },
  {
   "detail": "1 fork point(s) at seq [1987]: two pipeline processes appended to one file concurrently, each continuing from the tail it read at start-up. Every row hashes correctly and links to an existing row, so this is a concurrency defect, not tampering. gate_audit.json says chain_ok=True but checked only 12 in-process rows of 2094. The gate audit log is also not part of the signed bundle.",
   "effect": "Writer fixed (file lock + on-disk tail) in cain/cain2027/e80/core.py with a regression test; the historical file is preserved unmodified and must not be described as an unbroken chain.",
   "id": "F-02",
   "severity": "MEDIUM",
   "title": "Gate audit log is forked, and the summary reports chain_ok over a tail only"
  },
  {
   "detail": "rows account for 700 calls across 15 arms; the report says 412 (the budget of the process that finished the run after a resume); the ledger holds 855 rows for the mission, including calls from an interrupted first attempt whose arms were re-run.",
   "effect": "Accuracy figures are unaffected. Quote 700 calls for the scored rows, not 412.",
   "id": "F-03",
   "severity": "LOW",
   "title": "Benchmark call count reflects only the final resumed process"
  },
  {
   "detail": "preregistration.json created_utc=1794828568.809831 but the first M-bench call is at 1794825050.862719 (59 min earlier). The file is rewritten at every process start and the run was resumed, so the original timestamp was overwritten.",
   "effect": "The criteria are fixed in code and the hypothesis was REJECTED, so there is no selective-reporting gain, but 'written before the run' cannot be shown from the artifact. Writer fixed to never overwrite.",
   "id": "F-04",
   "severity": "MEDIUM",
   "title": "Pre-registration timestamp postdates the first benchmark calls"
  },
  {
   "detail": "`*.db` is gitignored, so the SQLite mission/endurance stores were never in the repository, and evolution-81 had not been pushed to the development replica when the host lost power (the replica held evolution-80 only). The stores survived on disk; a disk loss would have left only JSON exports of them.",
   "effect": "Stores force-added to the branch as evidence and the branch pushed to the replica by the takeover review. Replicate after every milestone commit, not at the end.",
   "id": "F-05",
   "severity": "MEDIUM",
   "title": "Durable stores and the branch itself existed on one host only"
  }
 ],
 "host_power_loss_note": "The host stopped logging at 2026-11-16T13:42:59Z and booted at 15:57Z. No E81 mission was running, so this shows durable state surviving a real power loss; it is not a mission resumed mid-flight across a reboot (G07 stays PARTIAL).",
 "imports_cain_code": false,
 "limits": [
  "verifies records, not the world",
  "same project and same administrative domain",
  "task ground truth is taken from the stored `correct` flags; it is not re-derived here"
 ],
 "ok": true,
 "recomputed": {
  "capacity": {
   "concurrent_model_inference_in_capacity_stages": 0,
   "max_real_worker_processes": 500,
   "not_run": [
    {
     "reason": "attempted twice; SQLite write contention at 1,000 concurrent claiming processes did not complete within the wall cap. 500 real workers is the measured maximum on this host.",
     "stage": 1000
    },
    {
     "reason": "logical identities only (records), never concurrent processes",
     "stage": 5000
    },
    {
     "reason": "logical identities only (records), never concurrent processes",
     "stage": 10000
    }
   ],
   "stages": [
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 9.759,
     "tasks_done": 40,
     "workers": 10
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 131.48,
     "tasks_done": 200,
     "workers": 50
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 434.674,
     "tasks_done": 400,
     "workers": 100
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 2146.473,
     "tasks_done": 1000,
     "workers": 250
    },
    {
     "concurrent_inference": 0,
     "duplicate_effects": 0,
     "failed": 0,
     "p95_ms": 6082.557,
     "tasks_done": 2000,
     "workers": 500
    }
   ]
  },
  "collaboration": {
   "arms": {
    "A_single": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 24,
     "losses_vs_single": 0,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 0
    },
    "B_hierarchical": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 48,
     "losses_vs_single": 3,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 3
    },
    "C_parallel_research": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 72,
     "losses_vs_single": 0,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 0
    },
    "D_evidence_first": {
     "accuracy": 0.2917,
     "correct": 7,
     "logical_calls": 24,
     "losses_vs_single": 8,
     "mcnemar_exact_p": 0.581055,
     "n": 24,
     "wilson95": [
      0.1491,
      0.4917
     ],
     "wins_vs_single": 5
    },
    "E_adversarial_review": {
     "accuracy": 0.3333,
     "correct": 8,
     "logical_calls": 48,
     "losses_vs_single": 3,
     "mcnemar_exact_p": 0.625,
     "n": 24,
     "wilson95": [
      0.1797,
      0.5329
     ],
     "wins_vs_single": 1
    },
    "F_debate_synthesis": {
     "accuracy": 0.3333,
     "correct": 8,
     "logical_calls": 72,
     "losses_vs_single": 2,
     "mcnemar_exact_p": 0.5,
     "n": 24,
     "wilson95": [
      0.1797,
      0.5329
     ],
     "wins_vs_single": 0
    },
    "G_shared_workspace": {
     "accuracy": 0.3333,
     "correct": 8,
     "logical_calls": 48,
     "losses_vs_single": 3,
     "mcnemar_exact_p": 0.625,
     "n": 24,
     "wilson95": [
      0.1797,
      0.5329
     ],
     "wins_vs_single": 1
    },
    "H_dynamic_routing": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 24,
     "losses_vs_single": 0,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 0
    },
    "I_fixed_routing": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 24,
     "losses_vs_single": 0,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 0
    },
    "J_confidence_aware": {
     "accuracy": 0.0,
     "correct": 0,
     "logical_calls": 41,
     "losses_vs_single": 10,
     "mcnemar_exact_p": 0.001953,
     "n": 24,
     "wilson95": [
      0.0,
      0.138
     ],
     "wins_vs_single": 0
    },
    "K_model_specialist": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 24,
     "losses_vs_single": 0,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 0
    },
    "L_adaptive_comm": {
     "accuracy": 0.375,
     "correct": 9,
     "logical_calls": 59,
     "losses_vs_single": 1,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2116,
      0.5729
     ],
     "wins_vs_single": 0
    },
    "M_independent_vote": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 72,
     "losses_vs_single": 1,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 1
    },
    "N_multistage_verify": {
     "accuracy": 0.3333,
     "correct": 8,
     "logical_calls": 48,
     "losses_vs_single": 3,
     "mcnemar_exact_p": 0.625,
     "n": 24,
     "wilson95": [
      0.1797,
      0.5329
     ],
     "wins_vs_single": 1
    },
    "O_cross_domain": {
     "accuracy": 0.4167,
     "correct": 10,
     "logical_calls": 72,
     "losses_vs_single": 3,
     "mcnemar_exact_p": 1.0,
     "n": 24,
     "wilson95": [
      0.2447,
      0.6117
     ],
     "wins_vs_single": 3
    }
   },
   "best_collaborative_arm": "B_hierarchical",
   "best_delta": 0.0,
   "calls_in_ledger_for_mission": 855,
   "calls_in_rows": 700,
   "calls_reported": 412,
   "collaboration_superior": false,
   "first_benchmark_call_utc": 1794825050.862719,
   "preregistration_created_utc": 1794828568.809831,
   "single_agent_accuracy": 0.4167
  },
  "durable_stores": [
   {
    "bytes": 176128,
    "duplicate_effect_keys": 0,
    "integrity_check": "ok",
    "rows": {
     "checkpoints": 114,
     "dead_letters": 0,
     "effects": 116,
     "events": 470,
     "meta": 1,
     "missions": 1,
     "sqlite_sequence": 1,
     "tasks": 116
    },
    "store": "endurance/endurance.db",
    "task_states": {
     "DONE": 116
    },
    "wal_bytes": 0
   },
   {
    "bytes": 65536,
    "duplicate_effect_keys": 0,
    "integrity_check": "ok",
    "rows": {
     "checkpoints": 1,
     "dead_letters": 0,
     "effects": 6,
     "events": 25,
     "meta": 1,
     "missions": 1,
     "sqlite_sequence": 1,
     "tasks": 6
    },
    "store": "work/mission/mission.db",
    "task_states": {
     "DONE": 6
    },
    "wal_bytes": null
   }
  ],
  "endurance": {
   "elapsed_s": 3600.0,
   "ledger_calls": 114
  },
  "gate_audit": {
   "decisions": {
    "DECIDE:ALLOW": 1047,
    "DECIDE:DENY": 2,
    "EXECUTE:EXECUTED": 1045
   },
   "forks": [
    {
     "children": 2,
     "parent_hash": "ecbe910341ff70f8",
     "seq": 1987
    }
   ],
   "linear_chain": false,
   "rows": 2094,
   "summary_file_chain_ok": true,
   "summary_file_events_checked": 12
  },
  "model_calls": {
   "by_execution_class": {
    "LOCAL_MODEL_INFERENCE": 319,
    "REAL_LOCAL_MODEL": 1156
   },
   "by_mission": {
    "M-bench": 855,
    "M-competency": 270,
    "M-memory": 64,
    "M-model-eval": 120,
    "M-reference-1": 9,
    "M-rep": 24,
    "M-routing": 132,
    "M-t": 1
   },
   "by_model": {
    "qwen2.5-0.5b-instruct": 1341,
    "qwen2.5-1.5b-instruct": 85,
    "smollm2-135m-instruct": 49
   },
   "endurance_ledger_rows": 114,
   "inference_wall_seconds": 5577.05,
   "input_tokens": 70743,
   "main_ledger_real_local": 1475,
   "main_ledger_rows": 1475,
   "output_tokens": 15818,
   "real_local_calls_all_ledgers": 1589,
   "reported_input_tokens": 70743,
   "reported_output_tokens": 15818,
   "reported_real_local_calls": 1156
  }
 },
 "schema": "cain2027.e81.takeover_review.v1",
 "written_by": "takeover review session, 2026-11-16; same project, not a third party"
}
