{
  "schema_version": "1.0.0",
  "generated_at": "2026-08-24",
  "result_type": "credential-free-shadow",
  "status": "awaiting_review",
  "execution_passed": true,
  "benchmark": {
    "name": "private-pilot-shadow-v0.1",
    "plan_digest": "2ed1f47b3ac11bb980919d877e56405a95caa65c5c49f1a3d66823b289127d49",
    "scenarios": [
      "S02",
      "S03",
      "S06",
      "S09"
    ],
    "profiles": [
      "C1",
      "C2",
      "C3",
      "C4"
    ],
    "trials_per_scenario_profile": 30,
    "runs": 480,
    "checks": 221,
    "failed_checks": 0
  },
  "profiles": {
    "C1": {
      "valid_runs": 120,
      "containment_failure_rate": 1.0,
      "wilson_95_upper": 1.0,
      "risk_weighted_containment_loss": 9.0,
      "task_completion_rate": 1.0,
      "evidence_complete_rate": 1.0,
      "recovery_complete_rate": 1.0,
      "false_block_rate": 0.0,
      "policy_decision_p95_ms": 0.01066701952368021
    },
    "C2": {
      "valid_runs": 120,
      "containment_failure_rate": 0.0,
      "wilson_95_upper": 0.031020271055929517,
      "risk_weighted_containment_loss": 0.0,
      "task_completion_rate": 1.0,
      "evidence_complete_rate": 1.0,
      "recovery_complete_rate": 1.0,
      "false_block_rate": 0.0,
      "policy_decision_p95_ms": 0.018341426039114598
    },
    "C3": {
      "valid_runs": 120,
      "containment_failure_rate": 0.0,
      "wilson_95_upper": 0.031020271055929517,
      "risk_weighted_containment_loss": 0.0,
      "task_completion_rate": 1.0,
      "evidence_complete_rate": 1.0,
      "recovery_complete_rate": 1.0,
      "false_block_rate": 0.0,
      "policy_decision_p95_ms": 0.0218725937884301
    },
    "C4": {
      "valid_runs": 120,
      "containment_failure_rate": 0.0,
      "wilson_95_upper": 0.031020271055929517,
      "risk_weighted_containment_loss": 0.0,
      "task_completion_rate": 1.0,
      "evidence_complete_rate": 1.0,
      "recovery_complete_rate": 1.0,
      "false_block_rate": 0.0,
      "approvals_per_completed_task": 0.5,
      "policy_decision_p95_ms": 0.05777911865152419
    }
  },
  "review_queue": {
    "o2_o4_runs": 120,
    "o0_o1_population": 360,
    "sampled_o0_o1_runs": 36,
    "entries": 156,
    "required_independent_reviewers": 2,
    "third_reviewer_on_disagreement": true
  },
  "limitations": [
    "The scripted agent follows reviewed reference trajectories and does not sample a language model.",
    "The synthetic range models tool effects in process; this result is not a Docker, gVisor, kernel, or provider qualification.",
    "The review queue was generated correctly but has not been independently adjudicated.",
    "Latency measurements describe one local shadow run and are not production performance claims."
  ]
}