{
  "benchmark": "synthetic_core_decision_contracts",
  "version": 1,
  "baseline_commit": "feb2bc21f34e5a7c918ae6dde8b35780c16e197a",
  "treatment_base_commit": "feb2bc21f34e5a7c918ae6dde8b35780c16e197a",
  "treatment_source_sha256": "44ad0ab2898de612d75660daef4b28178eef2344b2b63349f28d6bb9e393d45a",
  "benchmark_sha256": "7b0fc0c14d3c34a61c0f2e7fb459ff95f9aa3158312ff67ec0611af567a83a0c",
  "inputs": "13 fixed scripted synthetic distributions/failures in this module",
  "uncertainty_treatment": {
    "min_top_probability": 0.55,
    "min_margin": 0.1
  },
  "baseline": {
    "label": "baseline",
    "cases": [
      {
        "id": "requirement_clear",
        "expected": "MATCH",
        "actual": "MATCH",
        "correct": true,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 1756125,
        "request_context_bytes": 815,
        "returned_context_bytes": 74,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "requirement_weak",
        "expected": "REVIEW",
        "actual": "MATCH",
        "correct": false,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": true,
        "operation_elapsed_ns": 141750,
        "request_context_bytes": 815,
        "returned_context_bytes": 74,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "filter_clear_negative",
        "expected": "EXCLUDE",
        "actual": "EXCLUDE",
        "correct": true,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 126959,
        "request_context_bytes": 744,
        "returned_context_bytes": 70,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "filter_weak_negative",
        "expected": "REVIEW",
        "actual": "EXCLUDE",
        "correct": false,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": true,
        "operation_elapsed_ns": 102334,
        "request_context_bytes": 744,
        "returned_context_bytes": 70,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "choose_weak",
        "expected": "REVIEW",
        "actual": "MATCH",
        "correct": false,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": true,
        "operation_elapsed_ns": 89416,
        "request_context_bytes": 548,
        "returned_context_bytes": 74,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "hosted_clear_target",
        "expected": "done",
        "actual": "done",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 1,
        "executed_actions": 1,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 69542,
        "request_context_bytes": 10674,
        "returned_context_bytes": 570,
        "bookkeeping": {
          "usage": {
            "input_tokens": 200,
            "output_tokens": 20
          },
          "usage_complete": true,
          "verification": null
        }
      },
      {
        "id": "hosted_weak_target",
        "expected": "needs_review",
        "actual": "done",
        "correct": false,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 1,
        "false_actions": 1,
        "review": false,
        "unsafe_automatic": true,
        "operation_elapsed_ns": 39959,
        "request_context_bytes": 10674,
        "returned_context_bytes": 571,
        "bookkeeping": {
          "usage": {
            "input_tokens": 200,
            "output_tokens": 20
          },
          "usage_complete": true,
          "verification": null
        }
      },
      {
        "id": "hosted_weak_value",
        "expected": "needs_review",
        "actual": "done",
        "correct": false,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 1,
        "false_actions": 1,
        "review": false,
        "unsafe_automatic": true,
        "operation_elapsed_ns": 39125,
        "request_context_bytes": 16477,
        "returned_context_bytes": 635,
        "bookkeeping": {
          "usage": {
            "input_tokens": 200,
            "output_tokens": 20
          },
          "usage_complete": true,
          "verification": null
        }
      },
      {
        "id": "hosted_weak_operation",
        "expected": "needs_review",
        "actual": "done",
        "correct": false,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 1,
        "false_actions": 1,
        "review": false,
        "unsafe_automatic": true,
        "operation_elapsed_ns": 31417,
        "request_context_bytes": 10674,
        "returned_context_bytes": 586,
        "bookkeeping": {
          "usage": {
            "input_tokens": 200,
            "output_tokens": 20
          },
          "usage_complete": true,
          "verification": null
        }
      },
      {
        "id": "first_provider_failure",
        "expected": "error",
        "actual": "error",
        "correct": false,
        "bookkeeping_preserved": false,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 10583,
        "request_context_bytes": 5315,
        "returned_context_bytes": 344,
        "bookkeeping": {
          "usage": {
            "input_tokens": 0,
            "output_tokens": 0
          },
          "usage_complete": true,
          "error": "ProviderError",
          "verification": null
        }
      },
      {
        "id": "provider_failure_after_success",
        "expected": "error",
        "actual": "error",
        "correct": false,
        "bookkeeping_preserved": false,
        "expected_actions": 1,
        "executed_actions": 1,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 21583,
        "request_context_bytes": 10674,
        "returned_context_bytes": 574,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": true,
          "error": "ProviderError",
          "verification": null
        }
      },
      {
        "id": "verifier_failure",
        "expected": "unverified",
        "actual": "unverified",
        "correct": false,
        "bookkeeping_preserved": false,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 15292,
        "request_context_bytes": 5564,
        "returned_context_bytes": 412,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": true,
          "verification": {
            "passed": false,
            "question": false,
            "question_error": "ProviderError"
          }
        }
      },
      {
        "id": "text_helper_failure",
        "expected": "error",
        "actual": "raised",
        "correct": false,
        "bookkeeping_preserved": false,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 23708,
        "request_context_bytes": 7587,
        "returned_context_bytes": 50,
        "bookkeeping": {
          "error_type": "TextHelperError"
        }
      }
    ],
    "correct": 3,
    "case_count": 13,
    "unsafe_automatic_decisions": 6,
    "false_actions": 3,
    "reviews": 0,
    "operation_elapsed_ns": 2467793,
    "request_context_bytes": 81305,
    "returned_context_bytes": 4104,
    "actual_model_usage": {
      "requests": 0,
      "models": {},
      "input_tokens": 0,
      "output_tokens": 0,
      "usage_complete": true
    }
  },
  "treatment": {
    "label": "treatment",
    "cases": [
      {
        "id": "requirement_clear",
        "expected": "MATCH",
        "actual": "MATCH",
        "correct": true,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 1266167,
        "request_context_bytes": 815,
        "returned_context_bytes": 74,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "requirement_weak",
        "expected": "REVIEW",
        "actual": "REVIEW",
        "correct": true,
        "false_actions": 0,
        "review": true,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 152208,
        "request_context_bytes": 815,
        "returned_context_bytes": 288,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "filter_clear_negative",
        "expected": "EXCLUDE",
        "actual": "EXCLUDE",
        "correct": true,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 151250,
        "request_context_bytes": 744,
        "returned_context_bytes": 70,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "filter_weak_negative",
        "expected": "REVIEW",
        "actual": "REVIEW",
        "correct": true,
        "false_actions": 0,
        "review": true,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 139750,
        "request_context_bytes": 744,
        "returned_context_bytes": 293,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "choose_weak",
        "expected": "REVIEW",
        "actual": "REVIEW",
        "correct": true,
        "false_actions": 0,
        "review": true,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 115916,
        "request_context_bytes": 548,
        "returned_context_bytes": 287,
        "simulated_usage": {
          "input_tokens": 100,
          "output_tokens": 10
        }
      },
      {
        "id": "hosted_clear_target",
        "expected": "done",
        "actual": "done",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 1,
        "executed_actions": 1,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 91375,
        "request_context_bytes": 10674,
        "returned_context_bytes": 788,
        "bookkeeping": {
          "usage": {
            "input_tokens": 200,
            "output_tokens": 20
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "verification": null,
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "hosted_weak_target",
        "expected": "needs_review",
        "actual": "needs_review",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": true,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 36709,
        "request_context_bytes": 5315,
        "returned_context_bytes": 884,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "verification": null,
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "hosted_weak_value",
        "expected": "needs_review",
        "actual": "needs_review",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": true,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 37458,
        "request_context_bytes": 8207,
        "returned_context_bytes": 884,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "verification": null,
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "hosted_weak_operation",
        "expected": "needs_review",
        "actual": "needs_review",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": true,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 30458,
        "request_context_bytes": 5315,
        "returned_context_bytes": 779,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "verification": null,
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "first_provider_failure",
        "expected": "error",
        "actual": "error",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 21125,
        "request_context_bytes": 5315,
        "returned_context_bytes": 609,
        "bookkeeping": {
          "usage": {
            "input_tokens": 0,
            "output_tokens": 0
          },
          "usage_complete": false,
          "unknown_usage_attempts": 1,
          "error": "transport_failed_usage_unknown",
          "error_type": "ProviderError",
          "verification": null,
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "provider_failure_after_success",
        "expected": "error",
        "actual": "error",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 1,
        "executed_actions": 1,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 44458,
        "request_context_bytes": 10674,
        "returned_context_bytes": 839,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": false,
          "unknown_usage_attempts": 1,
          "error": "transport_failed_usage_unknown",
          "error_type": "ProviderError",
          "verification": null,
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "verifier_failure",
        "expected": "unverified",
        "actual": "unverified",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 29958,
        "request_context_bytes": 5564,
        "returned_context_bytes": 686,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": false,
          "unknown_usage_attempts": 1,
          "verification": {
            "passed": false,
            "question": false,
            "question_error": "response_invalid_usage_unknown",
            "question_error_type": "ProviderError"
          },
          "text_model_usage": {
            "requests": 0,
            "models": {},
            "usage_complete": true,
            "unknown_usage_attempts": 0
          }
        }
      },
      {
        "id": "text_helper_failure",
        "expected": "error",
        "actual": "error",
        "correct": true,
        "bookkeeping_preserved": true,
        "expected_actions": 0,
        "executed_actions": 0,
        "false_actions": 0,
        "review": false,
        "unsafe_automatic": false,
        "operation_elapsed_ns": 32250,
        "request_context_bytes": 7587,
        "returned_context_bytes": 606,
        "bookkeeping": {
          "usage": {
            "input_tokens": 100,
            "output_tokens": 10
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "error": "text_model_unreachable",
          "error_type": "TextHelperError",
          "verification": null,
          "text_model_usage": {
            "requests": 1,
            "models": {},
            "usage_complete": false,
            "unknown_usage_attempts": 1
          }
        }
      }
    ],
    "correct": 13,
    "case_count": 13,
    "unsafe_automatic_decisions": 0,
    "false_actions": 0,
    "reviews": 6,
    "operation_elapsed_ns": 2149082,
    "request_context_bytes": 62317,
    "returned_context_bytes": 7087,
    "actual_model_usage": {
      "requests": 0,
      "models": {},
      "input_tokens": 0,
      "output_tokens": 0,
      "usage_complete": true
    }
  },
  "harness_elapsed_ns": 332262917,
  "fee_basis": {
    "actual_usd": 0.0,
    "reason": "offline scripted responses; no model requests or credential reads"
  },
  "limitations": [
    "Behavioral regression A/B, not model quality, calibration or real website success measurement.",
    "Fixture usage envelopes are simulated; actual model usage is explicitly zero for both arms.",
    "Timing includes complete simulated analysis/hosted operations, not model/network/browser latency.",
    "The treatment deliberately increases review and decreases automatic completion on ambiguous inputs.",
    "No result cache is used. Contracts and thresholds do not authorize actions."
  ]
}
