{
  "schema_version": 1,
  "kind": "live-browser-pinned-ab",
  "created_at": "2026-10-04T07:48:02.898880+00:00",
  "commits": {
    "baseline": "39d0bdc21dd3b17de452f413181f422a4c84af18",
    "treatment": "working-tree-snapshot"
  },
  "source_sha256": {
    "baseline": "329682d8b5e60d9ea81fe6a552ddc11af952c65f4b322db6f9bf42db77a3e18f",
    "treatment": "6c95c1640536085028953f8a71721ca83e1cb1a6576e408f78b617ce14f0b3dd"
  },
  "repeats": 3,
  "tasks": [
    "search-filter",
    "buy-pauses-before-order",
    "contact-form-scroll",
    "shadow-dom",
    "same-origin-frame",
    "delete-is-gated",
    "already-satisfied-controls",
    "nested-scroll-target"
  ],
  "fixture_sha256": "3ad5004db9b7876bfc1dec5885207cdcc3611a33a8b38809804cfc16a7d3049a",
  "runtime": {
    "python": "3.14.2",
    "system": "Darwin"
  },
  "whole_benchmark_elapsed_ms": 128484.479,
  "timing_scope": "fixture setup/navigation + observe/decide/act + independent final checks; browser startup reported separately",
  "usage_scope": "actual provider-reported input/output tokens; failed or retried unknown usage is preserved",
  "pricing": {
    "usd_per_million_input": 0.042,
    "output_billed": false,
    "basis": "estimate using stated Jev price, not an invoice"
  },
  "scope": "synthetic loopback fixtures in new private Chromium processes; no customer/business execution",
  "summary": {
    "baseline": {
      "runs": 24,
      "passed": 15,
      "correct_completions": 12,
      "correct_gates": 3,
      "review_or_failure": 9,
      "statuses": {
        "needs_review": 6,
        "done": 12,
        "needs_confirmation": 3,
        "blocked": 3
      },
      "false_completions": 0,
      "wrong_actions": 0,
      "model_calls": 68,
      "provider_attempts": 68,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 171320,
        "output_tokens": 12012
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 171320,
          "output_tokens": 12012
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "whole_operation_total_ms": 43354.236,
      "whole_operation_median_ms": 1353.742,
      "request_bytes": 598140,
      "state_bytes": 64552,
      "estimated_input_usd": 0.00719544,
      "known_input_usd_lower_bound": 0.00719544
    },
    "treatment": {
      "runs": 24,
      "passed": 18,
      "correct_completions": 15,
      "correct_gates": 3,
      "review_or_failure": 6,
      "statuses": {
        "needs_review": 6,
        "done": 15,
        "needs_confirmation": 3
      },
      "false_completions": 0,
      "wrong_actions": 0,
      "model_calls": 81,
      "provider_attempts": 81,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 202290,
        "output_tokens": 12818
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 202290,
          "output_tokens": 12818
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "whole_operation_total_ms": 45952.722,
      "whole_operation_median_ms": 1530.398,
      "request_bytes": 731412,
      "state_bytes": 74340,
      "estimated_input_usd": 0.00849618,
      "known_input_usd_lower_bound": 0.00849618
    }
  },
  "runs": [
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 989.23,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.45,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.020000000000000018,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 2035,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "results_page": false,
        "exact_search": false,
        "category": false,
        "stock": false,
        "consent_rejected": true
      },
      "whole_operation_ms": 2170.291,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 13630,
        "output_tokens": 922
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 13630,
          "output_tokens": 922
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 46283,
      "state_bytes": 6564,
      "estimated_input_usd": 0.00057246,
      "known_input_usd_lower_bound": 0.00057246,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5968,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 946.792
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12903,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 404.177
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13711,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 250.252
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13701,
          "state_bytes": 2156,
          "usage": {
            "input_tokens": 4120,
            "output_tokens": 278
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 233.298
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 375.572,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "operation",
          "reason": "below_min_top_probability",
          "actual": 0.51,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.04999999999999999,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1543,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "results_page": false,
        "exact_search": false,
        "category": false,
        "stock": false,
        "consent_rejected": true
      },
      "whole_operation_ms": 1612.932,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10230,
        "output_tokens": 644
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10230,
          "output_tokens": 644
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 36712,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00042966,
      "known_input_usd_lower_bound": 0.00042966,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6794,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1821,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 803.356
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14555,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 4004,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 363.39
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15363,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4405,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 236.794
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 366.505,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.45,
          "threshold": 0.55
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1596,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1671.364,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 644
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 644
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32417,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5935,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 760.707
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12837,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 429.348
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13645,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 263.331
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 365.656,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.37,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.03999999999999998,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1654,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1724.382,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10230,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10230,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 36547,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00042966,
      "known_input_usd_lower_bound": 0.00042966,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6761,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1821,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 897.425
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14489,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 4004,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 372.468
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15297,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4405,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 239.071
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 360.506,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3339,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3412.901,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21378,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21378,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77772,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000897876,
      "known_input_usd_lower_bound": 0.000897876,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 9038,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2429,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 848.76
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 9263,
          "state_bytes": 614,
          "usage": {
            "input_tokens": 2521,
            "output_tokens": 194
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 235.798
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 9431,
          "state_bytes": 726,
          "usage": {
            "input_tokens": 2587,
            "output_tokens": 195
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 257.268
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12417,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3421,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 375.699
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12463,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3442,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 446.493
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12553,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3476,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 407.407
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12607,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3502,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 404.805
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 366.486,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3257,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3330.114,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 23178,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 23178,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 88097,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000973476,
      "known_input_usd_lower_bound": 0.000973476,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 10277,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2645,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 862.495
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 10502,
          "state_bytes": 614,
          "usage": {
            "input_tokens": 2737,
            "output_tokens": 194
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 388.273
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 10670,
          "state_bytes": 726,
          "usage": {
            "input_tokens": 2803,
            "output_tokens": 195
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 392.958
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14069,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3709,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 235.518
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14115,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3730,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 244.179
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14205,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3764,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 382.337
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14259,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3790,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 392.272
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 369.756,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1260,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1358.261,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3268,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3268,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11352,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000137256,
      "known_input_usd_lower_bound": 0.000137256,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5642,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1621,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 821.457
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5710,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1647,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 390.289
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 391.924,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1243,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1364.569,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3556,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3556,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 13004,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000149352,
      "known_input_usd_lower_bound": 0.000149352,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6468,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1765,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 844.382
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6536,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1791,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 343.573
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 363.767,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1149,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1256.694,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3294,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3294,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11508,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000138348,
      "known_input_usd_lower_bound": 0.000138348,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5720,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1633,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 860.189
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5788,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1661,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 230.565
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 359.864,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1253,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1358.321,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3582,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3582,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 13160,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000150444,
      "known_input_usd_lower_bound": 0.000150444,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6546,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1777,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 815.019
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6614,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1805,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 381.307
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 405.956,
      "passed": true,
      "correct_completion": false,
      "correct_gate": true,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_confirmation",
      "reason": "label_rule",
      "review_reasons": [],
      "pending_label": "Remove Red Runner",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 856,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": true,
        "exact_cart_preserved": true,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 943.773,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1709,
        "output_tokens": 109
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1709,
          "output_tokens": 109
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 5842,
      "state_bytes": 594,
      "estimated_input_usd": 7.1778e-05,
      "known_input_usd_lower_bound": 7.1778e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5842,
          "state_bytes": 594,
          "usage": {
            "input_tokens": 1709,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 851.04
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 396.565,
      "passed": true,
      "correct_completion": false,
      "correct_gate": true,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_confirmation",
      "reason": "label_rule",
      "review_reasons": [],
      "pending_label": "Remove Red Runner",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 854,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": true,
        "exact_cart_preserved": true,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 940.611,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1853,
        "output_tokens": 109
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1853,
          "output_tokens": 109
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 6668,
      "state_bytes": 594,
      "estimated_input_usd": 7.7826e-05,
      "known_input_usd_lower_bound": 7.7826e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6668,
          "state_bytes": 594,
          "usage": {
            "input_tokens": 1853,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 849.691
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 384.135,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1234,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1317.811,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3732,
        "output_tokens": 246
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3732,
          "output_tokens": 246
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 12600,
      "state_bytes": 1459,
      "estimated_input_usd": 0.000156744,
      "known_input_usd_lower_bound": 0.000156744,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6260,
          "state_bytes": 687,
          "usage": {
            "input_tokens": 1852,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 820.004
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6340,
          "state_bytes": 772,
          "usage": {
            "input_tokens": 1880,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 358.516
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 367.924,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1135,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1217.298,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 4020,
        "output_tokens": 246
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 4020,
          "output_tokens": 246
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 14252,
      "state_bytes": 1459,
      "estimated_input_usd": 0.00016884,
      "known_input_usd_lower_bound": 0.00016884,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 7086,
          "state_bytes": 687,
          "usage": {
            "input_tokens": 1996,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 837.604
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7166,
          "state_bytes": 772,
          "usage": {
            "input_tokens": 2024,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 247.686
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 379.286,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "blocked",
      "reason": "model_blocked",
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1241,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": false
      },
      "whole_operation_ms": 1333.563,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1959,
        "output_tokens": 159
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1959,
          "output_tokens": 159
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 6173,
      "state_bytes": 494,
      "estimated_input_usd": 8.2278e-05,
      "known_input_usd_lower_bound": 8.2278e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 3031,
          "state_bytes": 206,
          "usage": {
            "input_tokens": 957,
            "output_tokens": 75
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 815.586
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 3142,
          "state_bytes": 288,
          "usage": {
            "input_tokens": 1002,
            "output_tokens": 84
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 369.275
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 435.872,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3108,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": true
      },
      "whole_operation_ms": 3186.911,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10781,
        "output_tokens": 613
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10781,
          "output_tokens": 613
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 35364,
      "state_bytes": 5194,
      "estimated_input_usd": 0.000452802,
      "known_input_usd_lower_bound": 0.000452802,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 3909,
          "state_bytes": 507,
          "usage": {
            "input_tokens": 1226,
            "output_tokens": 75
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 822.443
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4197,
          "state_bytes": 574,
          "usage": {
            "input_tokens": 1354,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 374.617
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4281,
          "state_bytes": 660,
          "usage": {
            "input_tokens": 1389,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 354.151
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4356,
          "state_bytes": 743,
          "usage": {
            "input_tokens": 1415,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 372.839
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4447,
          "state_bytes": 840,
          "usage": {
            "input_tokens": 1447,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 420.407
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7049,
          "state_bytes": 894,
          "usage": {
            "input_tokens": 1960,
            "output_tokens": 103
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 267.429
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7125,
          "state_bytes": 976,
          "usage": {
            "input_tokens": 1990,
            "output_tokens": 103
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 240.141
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 396.31,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "operation",
          "reason": "below_min_top_probability",
          "actual": 0.52,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.08000000000000002,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1730,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "results_page": false,
        "exact_search": false,
        "category": false,
        "stock": false,
        "consent_rejected": true
      },
      "whole_operation_ms": 1800.986,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10230,
        "output_tokens": 644
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10230,
          "output_tokens": 644
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 36712,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00042966,
      "known_input_usd_lower_bound": 0.00042966,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6794,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1821,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 908.249
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14555,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 4004,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 395.446
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15363,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4405,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 284.359
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 373.374,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.49,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.08999999999999997,
          "threshold": 0.1
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 1980,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "results_page": false,
        "exact_search": false,
        "category": false,
        "stock": false,
        "consent_rejected": true
      },
      "whole_operation_ms": 2055.335,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 13630,
        "output_tokens": 922
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 13630,
          "output_tokens": 922
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 46283,
      "state_bytes": 6564,
      "estimated_input_usd": 0.00057246,
      "known_input_usd_lower_bound": 0.00057246,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5968,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 888.028
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12903,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 361.765
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13711,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 246.399
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13701,
          "state_bytes": 2156,
          "usage": {
            "input_tokens": 4120,
            "output_tokens": 278
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 313.825
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 381.694,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.41000000000000003,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.09000000000000002,
          "threshold": 0.1
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1813,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1884.908,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10230,
        "output_tokens": 644
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10230,
          "output_tokens": 644
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 36547,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00042966,
      "known_input_usd_lower_bound": 0.00042966,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6761,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1821,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 835.779
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14489,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 4004,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 466.731
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15297,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4405,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 367.754
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 356.201,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.38,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.02999999999999997,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1669,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1747.092,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32417,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5935,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 895.119
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12837,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 366.619
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13645,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 274.264
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 370.417,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3237,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3318.807,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 23178,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 23178,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 88097,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000973476,
      "known_input_usd_lower_bound": 0.000973476,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 10277,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2645,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 910.571
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 10502,
          "state_bytes": 614,
          "usage": {
            "input_tokens": 2737,
            "output_tokens": 194
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 353.487
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 10670,
          "state_bytes": 726,
          "usage": {
            "input_tokens": 2803,
            "output_tokens": 195
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 396.938
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14069,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3709,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 238.171
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14115,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3730,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 258.633
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14205,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3764,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 360.463
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14259,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3790,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 350.78
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 364.105,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 6683,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 6752.649,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21378,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21378,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77772,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000897876,
      "known_input_usd_lower_bound": 0.000897876,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 9038,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2429,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 3864.84
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 9263,
          "state_bytes": 614,
          "usage": {
            "input_tokens": 2521,
            "output_tokens": 194
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 237.331
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 9431,
          "state_bytes": 726,
          "usage": {
            "input_tokens": 2587,
            "output_tokens": 195
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 360.121
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12417,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3421,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 385.054
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12463,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3442,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 399.568
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12553,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3476,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 665.951
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12607,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3502,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 406.849
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 382.993,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1343,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1447.864,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3556,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3556,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 13004,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000149352,
      "known_input_usd_lower_bound": 0.000149352,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6468,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1765,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 932.277
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6536,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1791,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 358.408
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 367.832,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1242,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1345.404,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3268,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3268,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11352,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000137256,
      "known_input_usd_lower_bound": 0.000137256,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5642,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1621,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 846.49
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5710,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1647,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 341.967
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 372.395,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1274,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1373.073,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3582,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3582,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 13160,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000150444,
      "known_input_usd_lower_bound": 0.000150444,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6546,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1777,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 845.52
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6614,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1805,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 368.255
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 369.876,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1250,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1346.71,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3294,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3294,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11508,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000138348,
      "known_input_usd_lower_bound": 0.000138348,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5720,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1633,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 830.651
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5788,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1661,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 362.42
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 361.263,
      "passed": true,
      "correct_completion": false,
      "correct_gate": true,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_confirmation",
      "reason": "label_rule",
      "review_reasons": [],
      "pending_label": "Remove Red Runner",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 868,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": true,
        "exact_cart_preserved": true,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 944.062,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1853,
        "output_tokens": 109
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1853,
          "output_tokens": 109
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 6668,
      "state_bytes": 594,
      "estimated_input_usd": 7.7826e-05,
      "known_input_usd_lower_bound": 7.7826e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6668,
          "state_bytes": 594,
          "usage": {
            "input_tokens": 1853,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 862.719
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 486.648,
      "passed": true,
      "correct_completion": false,
      "correct_gate": true,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_confirmation",
      "reason": "label_rule",
      "review_reasons": [],
      "pending_label": "Remove Red Runner",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 760,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": true,
        "exact_cart_preserved": true,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 836.334,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1709,
        "output_tokens": 109
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1709,
          "output_tokens": 109
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 5842,
      "state_bytes": 594,
      "estimated_input_usd": 7.1778e-05,
      "known_input_usd_lower_bound": 7.1778e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5842,
          "state_bytes": 594,
          "usage": {
            "input_tokens": 1709,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 755.882
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 378.376,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1197,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1282.2,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 4020,
        "output_tokens": 246
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 4020,
          "output_tokens": 246
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 14252,
      "state_bytes": 1459,
      "estimated_input_usd": 0.00016884,
      "known_input_usd_lower_bound": 0.00016884,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 7086,
          "state_bytes": 687,
          "usage": {
            "input_tokens": 1996,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 884.011
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7166,
          "state_bytes": 772,
          "usage": {
            "input_tokens": 2024,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 255.148
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 361.933,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1273,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1349.224,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3732,
        "output_tokens": 246
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3732,
          "output_tokens": 246
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 12600,
      "state_bytes": 1459,
      "estimated_input_usd": 0.000156744,
      "known_input_usd_lower_bound": 0.000156744,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6260,
          "state_bytes": 687,
          "usage": {
            "input_tokens": 1852,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 853.895
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6340,
          "state_bytes": 772,
          "usage": {
            "input_tokens": 1880,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 354.771
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 372.162,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3304,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": true
      },
      "whole_operation_ms": 3371.572,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10781,
        "output_tokens": 613
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10781,
          "output_tokens": 613
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 35364,
      "state_bytes": 5194,
      "estimated_input_usd": 0.000452802,
      "known_input_usd_lower_bound": 0.000452802,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 3909,
          "state_bytes": 507,
          "usage": {
            "input_tokens": 1226,
            "output_tokens": 75
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 761.571
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4197,
          "state_bytes": 574,
          "usage": {
            "input_tokens": 1354,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 398.362
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4281,
          "state_bytes": 660,
          "usage": {
            "input_tokens": 1389,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 375.229
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4356,
          "state_bytes": 743,
          "usage": {
            "input_tokens": 1415,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 455.695
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4447,
          "state_bytes": 840,
          "usage": {
            "input_tokens": 1447,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 403.981
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7049,
          "state_bytes": 894,
          "usage": {
            "input_tokens": 1960,
            "output_tokens": 103
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 392.548
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7125,
          "state_bytes": 976,
          "usage": {
            "input_tokens": 1990,
            "output_tokens": 103
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 257.159
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 364.749,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "blocked",
      "reason": "model_blocked",
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1263,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": false
      },
      "whole_operation_ms": 1346.868,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1959,
        "output_tokens": 159
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1959,
          "output_tokens": 159
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 6173,
      "state_bytes": 494,
      "estimated_input_usd": 8.2278e-05,
      "known_input_usd_lower_bound": 8.2278e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 3031,
          "state_bytes": 206,
          "usage": {
            "input_tokens": 957,
            "output_tokens": 75
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 848.22
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 3142,
          "state_bytes": 288,
          "usage": {
            "input_tokens": 1002,
            "output_tokens": 84
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 359.966
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 375.516,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "operation",
          "reason": "below_min_top_probability",
          "actual": 0.52,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.07,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1607,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "results_page": false,
        "exact_search": false,
        "category": false,
        "stock": false,
        "consent_rejected": true
      },
      "whole_operation_ms": 1684.863,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 644
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 644
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32582,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5968,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 762.985
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12903,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 389.163
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13711,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 309.503
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 364.411,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.44,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.030000000000000027,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1788,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "results_page": false,
        "exact_search": false,
        "category": false,
        "stock": false,
        "consent_rejected": true
      },
      "whole_operation_ms": 1857.867,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10230,
        "output_tokens": 644
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10230,
          "output_tokens": 644
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 36712,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00042966,
      "known_input_usd_lower_bound": 0.00042966,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6794,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1821,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 863.709
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14555,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 4004,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 348.785
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15363,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4405,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 421.132
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 361.385,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.52,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1515,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1588.002,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32417,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5935,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 738.887
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12837,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 404.616
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13645,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 232.791
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 374.483,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.38,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.02999999999999997,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 2043,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 2126.216,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10230,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10230,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 36547,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00042966,
      "known_input_usd_lower_bound": 0.00042966,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6761,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1821,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 1140.412
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14489,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 4004,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 378.867
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15297,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4405,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 390.948
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 360.755,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3665,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3740.241,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21378,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21378,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77772,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000897876,
      "known_input_usd_lower_bound": 0.000897876,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 9038,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2429,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 865.062
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 9263,
          "state_bytes": 614,
          "usage": {
            "input_tokens": 2521,
            "output_tokens": 194
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 362.875
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 9431,
          "state_bytes": 726,
          "usage": {
            "input_tokens": 2587,
            "output_tokens": 195
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 375.566
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12417,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3421,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 390.966
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12463,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3442,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 397.788
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12553,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3476,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 434.501
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12607,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3502,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 446.967
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 409.304,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3464,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3551.821,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 23178,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 23178,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 88097,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000973476,
      "known_input_usd_lower_bound": 0.000973476,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 10277,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2645,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 912.864
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 10502,
          "state_bytes": 614,
          "usage": {
            "input_tokens": 2737,
            "output_tokens": 194
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 362.373
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 10670,
          "state_bytes": 726,
          "usage": {
            "input_tokens": 2803,
            "output_tokens": 195
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 368.247
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14069,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3709,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 547.527
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14115,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3730,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 230.106
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14205,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3764,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 246.745
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 14259,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3790,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 390.113
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 384.333,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1264,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1366.99,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3268,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3268,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11352,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000137256,
      "known_input_usd_lower_bound": 0.000137256,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5642,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1621,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 830.458
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5710,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1647,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 376.473
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 356.742,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1254,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1352.366,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3556,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3556,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 13004,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000149352,
      "known_input_usd_lower_bound": 0.000149352,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6468,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1765,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 855.495
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6536,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1791,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 347.274
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 366.827,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1170,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1262.959,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3294,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3294,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11508,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000138348,
      "known_input_usd_lower_bound": 0.000138348,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5720,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1633,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 894.542
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5788,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1661,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 224.963
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 375.39,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1290,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1392.95,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3582,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3582,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 13160,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000150444,
      "known_input_usd_lower_bound": 0.000150444,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6546,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1777,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 855.289
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6614,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1805,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 379.799
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 365.811,
      "passed": true,
      "correct_completion": false,
      "correct_gate": true,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_confirmation",
      "reason": "label_rule",
      "review_reasons": [],
      "pending_label": "Remove Red Runner",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 720,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": true,
        "exact_cart_preserved": true,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 797.043,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1709,
        "output_tokens": 109
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1709,
          "output_tokens": 109
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 5842,
      "state_bytes": 594,
      "estimated_input_usd": 7.1778e-05,
      "known_input_usd_lower_bound": 7.1778e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5842,
          "state_bytes": 594,
          "usage": {
            "input_tokens": 1709,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 714.553
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 360.438,
      "passed": true,
      "correct_completion": false,
      "correct_gate": true,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_confirmation",
      "reason": "label_rule",
      "review_reasons": [],
      "pending_label": "Remove Red Runner",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 949,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_pending_action": true,
        "exact_cart_preserved": true,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1028.379,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1853,
        "output_tokens": 109
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1853,
          "output_tokens": 109
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 6668,
      "state_bytes": 594,
      "estimated_input_usd": 7.7826e-05,
      "known_input_usd_lower_bound": 7.7826e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6668,
          "state_bytes": 594,
          "usage": {
            "input_tokens": 1853,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 942.916
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 359.52,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1284,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1359.236,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3732,
        "output_tokens": 246
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3732,
          "output_tokens": 246
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 12600,
      "state_bytes": 1459,
      "estimated_input_usd": 0.000156744,
      "known_input_usd_lower_bound": 0.000156744,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6260,
          "state_bytes": 687,
          "usage": {
            "input_tokens": 1852,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 884.722
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6340,
          "state_bytes": 772,
          "usage": {
            "input_tokens": 1880,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 346.609
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 371.005,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1157,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1229.711,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 4020,
        "output_tokens": 246
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 4020,
          "output_tokens": 246
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 14252,
      "state_bytes": 1459,
      "estimated_input_usd": 0.00016884,
      "known_input_usd_lower_bound": 0.00016884,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 7086,
          "state_bytes": 687,
          "usage": {
            "input_tokens": 1996,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 862.479
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7166,
          "state_bytes": 772,
          "usage": {
            "input_tokens": 2024,
            "output_tokens": 123
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 241.517
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 365.904,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "blocked",
      "reason": "model_blocked",
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1234,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": false
      },
      "whole_operation_ms": 1310.628,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 1959,
        "output_tokens": 159
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 1959,
          "output_tokens": 159
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 6173,
      "state_bytes": 494,
      "estimated_input_usd": 8.2278e-05,
      "known_input_usd_lower_bound": 8.2278e-05,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 3031,
          "state_bytes": 206,
          "usage": {
            "input_tokens": 957,
            "output_tokens": 75
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 837.261
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 3142,
          "state_bytes": 288,
          "usage": {
            "input_tokens": 1002,
            "output_tokens": 84
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 354.456
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 366.787,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 6,
      "kernel_elapsed_ms": 3182,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": true
      },
      "whole_operation_ms": 3254.802,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 10781,
        "output_tokens": 613
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 10781,
          "output_tokens": 613
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 35364,
      "state_bytes": 5194,
      "estimated_input_usd": 0.000452802,
      "known_input_usd_lower_bound": 0.000452802,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 3909,
          "state_bytes": 507,
          "usage": {
            "input_tokens": 1226,
            "output_tokens": 75
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 835.787
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4197,
          "state_bytes": 574,
          "usage": {
            "input_tokens": 1354,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 386.187
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4281,
          "state_bytes": 660,
          "usage": {
            "input_tokens": 1389,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 396.262
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4356,
          "state_bytes": 743,
          "usage": {
            "input_tokens": 1415,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 376.273
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 4447,
          "state_bytes": 840,
          "usage": {
            "input_tokens": 1447,
            "output_tokens": 83
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 355.047
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7049,
          "state_bytes": 894,
          "usage": {
            "input_tokens": 1960,
            "output_tokens": 103
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 363.47
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 7125,
          "state_bytes": 976,
          "usage": {
            "input_tokens": 1990,
            "output_tokens": 103
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 231.71
        }
      ],
      "arm": "treatment",
      "repeat": 3
    }
  ],
  "limitations": [
    "Small fixture sets and repeated runs cannot establish production reliability or calibrated accuracy.",
    "This compares released program versions on the same Jev model, not Jev against an LLM.",
    "Needed confirmation is a correct expected stop; it is never resumed in the benchmark.",
    "Synthetic final-state checks reject a false DONE but do not prove every intermediate action was optimal."
  ],
  "comparisons": {
    "existing-six": {
      "baseline": {
        "runs": 18,
        "passed": 12,
        "correct_completions": 9,
        "correct_gates": 3,
        "review_or_failure": 6,
        "statuses": {
          "needs_review": 6,
          "done": 9,
          "needs_confirmation": 3
        },
        "false_completions": 0,
        "wrong_actions": 0,
        "model_calls": 56,
        "provider_attempts": 56,
        "provider_retries": 0,
        "failed_provider_calls": 0,
        "known_usage": {
          "input_tokens": 154247,
          "output_tokens": 10797
        },
        "model_usage": {
          "jev-1.13.0": {
            "input_tokens": 154247,
            "output_tokens": 10797
          }
        },
        "usage_complete": true,
        "unknown_usage_attempts": 0,
        "whole_operation_total_ms": 35336.906,
        "whole_operation_median_ms": 1477.496,
        "request_bytes": 541821,
        "state_bytes": 58693,
        "estimated_input_usd": 0.006478374,
        "known_input_usd_lower_bound": 0.006478374
      },
      "treatment": {
        "runs": 18,
        "passed": 12,
        "correct_completions": 9,
        "correct_gates": 3,
        "review_or_failure": 6,
        "statuses": {
          "needs_review": 6,
          "done": 9,
          "needs_confirmation": 3
        },
        "false_completions": 0,
        "wrong_actions": 0,
        "model_calls": 54,
        "provider_attempts": 54,
        "provider_retries": 0,
        "failed_provider_calls": 0,
        "known_usage": {
          "input_tokens": 157887,
          "output_tokens": 10241
        },
        "model_usage": {
          "jev-1.13.0": {
            "input_tokens": 157887,
            "output_tokens": 10241
          }
        },
        "usage_complete": true,
        "unknown_usage_attempts": 0,
        "whole_operation_total_ms": 32410.228,
        "whole_operation_median_ms": 1530.398,
        "request_bytes": 582564,
        "state_bytes": 54381,
        "estimated_input_usd": 0.006631254,
        "known_input_usd_lower_bound": 0.006631254
      }
    },
    "nested-scroll": {
      "baseline": {
        "runs": 3,
        "passed": 0,
        "correct_completions": 0,
        "correct_gates": 0,
        "review_or_failure": 3,
        "statuses": {
          "blocked": 3
        },
        "false_completions": 0,
        "wrong_actions": 0,
        "model_calls": 6,
        "provider_attempts": 6,
        "provider_retries": 0,
        "failed_provider_calls": 0,
        "known_usage": {
          "input_tokens": 5877,
          "output_tokens": 477
        },
        "model_usage": {
          "jev-1.13.0": {
            "input_tokens": 5877,
            "output_tokens": 477
          }
        },
        "usage_complete": true,
        "unknown_usage_attempts": 0,
        "whole_operation_total_ms": 3991.059,
        "whole_operation_median_ms": 1333.563,
        "request_bytes": 18519,
        "state_bytes": 1482,
        "estimated_input_usd": 0.000246834,
        "known_input_usd_lower_bound": 0.000246834
      },
      "treatment": {
        "runs": 3,
        "passed": 3,
        "correct_completions": 3,
        "correct_gates": 0,
        "review_or_failure": 0,
        "statuses": {
          "done": 3
        },
        "false_completions": 0,
        "wrong_actions": 0,
        "model_calls": 21,
        "provider_attempts": 21,
        "provider_retries": 0,
        "failed_provider_calls": 0,
        "known_usage": {
          "input_tokens": 32343,
          "output_tokens": 1839
        },
        "model_usage": {
          "jev-1.13.0": {
            "input_tokens": 32343,
            "output_tokens": 1839
          }
        },
        "usage_complete": true,
        "unknown_usage_attempts": 0,
        "whole_operation_total_ms": 9813.285,
        "whole_operation_median_ms": 3254.802,
        "request_bytes": 106092,
        "state_bytes": 15582,
        "estimated_input_usd": 0.001358406,
        "known_input_usd_lower_bound": 0.001358406
      }
    },
    "already-satisfied-controls": {
      "baseline": {
        "runs": 3,
        "passed": 3,
        "correct_completions": 3,
        "correct_gates": 0,
        "review_or_failure": 0,
        "statuses": {
          "done": 3
        },
        "false_completions": 0,
        "wrong_actions": 0,
        "model_calls": 6,
        "provider_attempts": 6,
        "provider_retries": 0,
        "failed_provider_calls": 0,
        "known_usage": {
          "input_tokens": 11196,
          "output_tokens": 738
        },
        "model_usage": {
          "jev-1.13.0": {
            "input_tokens": 11196,
            "output_tokens": 738
          }
        },
        "usage_complete": true,
        "unknown_usage_attempts": 0,
        "whole_operation_total_ms": 4026.271,
        "whole_operation_median_ms": 1349.224,
        "request_bytes": 37800,
        "state_bytes": 4377,
        "estimated_input_usd": 0.000470232,
        "known_input_usd_lower_bound": 0.000470232
      },
      "treatment": {
        "runs": 3,
        "passed": 3,
        "correct_completions": 3,
        "correct_gates": 0,
        "review_or_failure": 0,
        "statuses": {
          "done": 3
        },
        "false_completions": 0,
        "wrong_actions": 0,
        "model_calls": 6,
        "provider_attempts": 6,
        "provider_retries": 0,
        "failed_provider_calls": 0,
        "known_usage": {
          "input_tokens": 12060,
          "output_tokens": 738
        },
        "model_usage": {
          "jev-1.13.0": {
            "input_tokens": 12060,
            "output_tokens": 738
          }
        },
        "usage_complete": true,
        "unknown_usage_attempts": 0,
        "whole_operation_total_ms": 3729.209,
        "whole_operation_median_ms": 1229.711,
        "request_bytes": 42756,
        "state_bytes": 4377,
        "estimated_input_usd": 0.00050652,
        "known_input_usd_lower_bound": 0.00050652
      }
    }
  }
}
