{
  "schema_version": 1,
  "kind": "live-browser-pinned-ab",
  "created_at": "2026-10-04T07:45:38.431396+00:00",
  "commits": {
    "baseline": "working-tree-snapshot",
    "treatment": "working-tree-snapshot"
  },
  "source_sha256": {
    "baseline": "4a73f6d212df8f71a41348902558015627935bfd2f972e240ff2f69be8d449d0",
    "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": 123168.401,
  "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": 14,
      "correct_completions": 11,
      "correct_gates": 3,
      "review_or_failure": 10,
      "statuses": {
        "needs_review": 9,
        "done": 11,
        "needs_confirmation": 3,
        "blocked": 1
      },
      "false_completions": 0,
      "wrong_actions": 0,
      "model_calls": 63,
      "provider_attempts": 63,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 156828,
        "output_tokens": 10873
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 156828,
          "output_tokens": 10873
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "whole_operation_total_ms": 38395.255,
      "whole_operation_median_ms": 1336.557,
      "request_bytes": 550159,
      "state_bytes": 59081,
      "estimated_input_usd": 0.006586776
    },
    "treatment": {
      "runs": 24,
      "passed": 18,
      "correct_completions": 15,
      "correct_gates": 3,
      "review_or_failure": 6,
      "statuses": {
        "needs_review": 5,
        "error": 1,
        "done": 15,
        "needs_confirmation": 3
      },
      "false_completions": 0,
      "wrong_actions": 0,
      "model_calls": 82,
      "provider_attempts": 82,
      "provider_retries": 0,
      "failed_provider_calls": 1,
      "known_usage": {
        "input_tokens": 202293,
        "output_tokens": 12809
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 202293,
          "output_tokens": 12809
        }
      },
      "usage_complete": false,
      "unknown_usage_attempts": 1,
      "whole_operation_total_ms": 46123.28,
      "whole_operation_median_ms": 1559.994,
      "request_bytes": 746765,
      "state_bytes": 76496,
      "estimated_input_usd": null
    }
  },
  "runs": [
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 940.541,
      "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": 0.51,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1796,
      "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": 1869.347,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9404,
        "output_tokens": 631
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9404,
          "output_tokens": 631
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32348,
      "state_bytes": 4390,
      "estimated_input_usd": 0.000394968,
      "known_input_usd_lower_bound": 0.000394968,
      "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": 863.679
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12786,
          "state_bytes": 1603,
          "usage": {
            "input_tokens": 3663,
            "output_tokens": 249
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 377.847
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13594,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 388.125
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 356.939,
      "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": 0.54,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.51,
          "threshold": 0.55
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1740,
      "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": 1815.691,
      "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": 857.462
        },
        {
          "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": 373.637
        },
        {
          "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": 356.393
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 365.247,
      "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": 0.39,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.04999999999999999,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1542,
      "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": 1615.116,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9404,
        "output_tokens": 631
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9404,
          "output_tokens": 631
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32183,
      "state_bytes": 4390,
      "estimated_input_usd": 0.000394968,
      "known_input_usd_lower_bound": 0.000394968,
      "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": 745.818
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12720,
          "state_bytes": 1603,
          "usage": {
            "input_tokens": 3663,
            "output_tokens": 249
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 378.582
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13528,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 277.848
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 369.041,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "error",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": "ProviderError",
      "error": "response_invalid_usage_unknown",
      "steps": 2,
      "kernel_elapsed_ms": 1654,
      "kernel_usage_complete": false,
      "kernel_unknown_usage_attempts": 1,
      "independent_checks": {
        "expected_pending_action": false,
        "exact_cart_preserved": false,
        "no_order_confirmation": true
      },
      "whole_operation_ms": 1735.236,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 1,
      "known_usage": {
        "input_tokens": 5825,
        "output_tokens": 358
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 5825,
          "output_tokens": 358
        }
      },
      "usage_complete": false,
      "unknown_usage_attempts": 1,
      "request_bytes": 36547,
      "state_bytes": 4408,
      "estimated_input_usd": null,
      "known_input_usd_lower_bound": 0.00024465,
      "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": 889.956
        },
        {
          "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": 382.482
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15297,
          "state_bytes": 2093,
          "error_type": "ProviderError",
          "usage_complete": false,
          "unknown_usage_attempts": 1,
          "attempts": 1,
          "elapsed_ms": 243.674
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 371.403,
      "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": 0.49,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.03999999999999998,
          "threshold": 0.1
        }
      ],
      "pending_label": "I agree to be contacted *",
      "error_type": null,
      "error": null,
      "steps": 4,
      "kernel_elapsed_ms": 2609,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": false,
        "exact_form_values": false
      },
      "whole_operation_ms": 2711.593,
      "model_calls": 5,
      "provider_attempts": 5,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 14296,
        "output_tokens": 1063
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 14296,
          "output_tokens": 1063
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 52411,
      "state_bytes": 4078,
      "estimated_input_usd": 0.000600432,
      "known_input_usd_lower_bound": 0.000600432,
      "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": 833.416
        },
        {
          "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": 398.163
        },
        {
          "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": 363.188
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12313,
          "state_bytes": 1080,
          "usage": {
            "input_tokens": 3368,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 386.221
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12366,
          "state_bytes": 1153,
          "usage": {
            "input_tokens": 3391,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 358.465
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 363.273,
      "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": 3387,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3465.007,
      "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": 848.353
        },
        {
          "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": 355.197
        },
        {
          "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": 380.385
        },
        {
          "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": 539.12
        },
        {
          "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": 279.109
        },
        {
          "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": 230.952
        },
        {
          "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": 353.636
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 393.571,
      "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": 1151,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1254.852,
      "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": 869.398
        },
        {
          "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": 229.23
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 373.498,
      "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": 1282,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1391.839,
      "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": 853.051
        },
        {
          "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": 390.506
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 367.953,
      "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": 1026,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1123.217,
      "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": 731.782
        },
        {
          "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": 238.547
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 357.15,
      "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": 1279,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1384.916,
      "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": 835.025
        },
        {
          "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": 389.964
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 428.758,
      "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": 852,
      "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": 938.442,
      "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": 848.393
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 366.653,
      "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": 867,
      "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": 943.523,
      "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": 860.202
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 363.543,
      "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": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1347.52,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3662,
        "output_tokens": 232
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3662,
          "output_tokens": 232
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 12472,
      "state_bytes": 1475,
      "estimated_input_usd": 0.000153804,
      "known_input_usd_lower_bound": 0.000153804,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6196,
          "state_bytes": 695,
          "usage": {
            "input_tokens": 1817,
            "output_tokens": 116
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 856.989
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6276,
          "state_bytes": 780,
          "usage": {
            "input_tokens": 1845,
            "output_tokens": 116
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 360.451
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 395.84,
      "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": 1267,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1352.883,
      "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": 829.826
        },
        {
          "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": 380.954
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 360.189,
      "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": 0.54,
          "threshold": 0.55
        }
      ],
      "pending_label": "Scroll down in Results panel",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 867,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": false
      },
      "whole_operation_ms": 951.933,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 957,
        "output_tokens": 75
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 957,
          "output_tokens": 75
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 3031,
      "state_bytes": 206,
      "estimated_input_usd": 4.0194e-05,
      "known_input_usd_lower_bound": 4.0194e-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": 863.136
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 375.187,
      "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": 3287,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": true
      },
      "whole_operation_ms": 3366.298,
      "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": 852.44
        },
        {
          "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": 353.409
        },
        {
          "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": 343.773
        },
        {
          "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": 380.305
        },
        {
          "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": 395.103
        },
        {
          "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": 364.516
        },
        {
          "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": 352.756
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 358.453,
      "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": 0.010000000000000009,
          "threshold": 0.1
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 2119,
      "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": 2199.851,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 14638,
        "output_tokens": 922
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 14638,
          "output_tokens": 922
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 52065,
      "state_bytes": 6564,
      "estimated_input_usd": 0.000614796,
      "known_input_usd_lower_bound": 0.000614796,
      "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": 884.218
        },
        {
          "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": 441.507
        },
        {
          "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": 239.926
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 15353,
          "state_bytes": 2156,
          "usage": {
            "input_tokens": 4408,
            "output_tokens": 278
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 360.485
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 365.089,
      "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
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 1936,
      "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": 2004.884,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 13496,
        "output_tokens": 911
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 13496,
          "output_tokens": 911
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 45992,
      "state_bytes": 6548,
      "estimated_input_usd": 0.000566832,
      "known_input_usd_lower_bound": 0.000566832,
      "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": 863.025
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12786,
          "state_bytes": 1603,
          "usage": {
            "input_tokens": 3663,
            "output_tokens": 249
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 392.86
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13594,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 263.106
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13644,
          "state_bytes": 2158,
          "usage": {
            "input_tokens": 4092,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 239.628
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 360.905,
      "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": 0.46,
          "threshold": 0.55
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1715,
      "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": 1787.577,
      "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": 842.689
        },
        {
          "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": 367.897
        },
        {
          "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": 351.057
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 387.585,
      "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": 0.51,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1731,
      "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": 1805.428,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9404,
        "output_tokens": 631
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9404,
          "output_tokens": 631
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32183,
      "state_bytes": 4390,
      "estimated_input_usd": 0.000394968,
      "known_input_usd_lower_bound": 0.000394968,
      "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": 834.349
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12720,
          "state_bytes": 1603,
          "usage": {
            "input_tokens": 3663,
            "output_tokens": 249
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 375.314
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13528,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 368.011
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 368.703,
      "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": 3426,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3498.494,
      "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": 878.321
        },
        {
          "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": 364.452
        },
        {
          "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": 400.127
        },
        {
          "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": 240.763
        },
        {
          "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": 388.907
        },
        {
          "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": 397.292
        },
        {
          "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": 390.576
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 362.353,
      "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": 3497,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3570.761,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21172,
        "output_tokens": 1552
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21172,
          "output_tokens": 1552
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77377,
      "state_bytes": 6651,
      "estimated_input_usd": 0.000889224,
      "known_input_usd_lower_bound": 0.000889224,
      "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": 889.897
        },
        {
          "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": 281.749
        },
        {
          "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": 367.377
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12313,
          "state_bytes": 1080,
          "usage": {
            "input_tokens": 3368,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 415.338
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12366,
          "state_bytes": 1153,
          "usage": {
            "input_tokens": 3391,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 421.617
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12456,
          "state_bytes": 1232,
          "usage": {
            "input_tokens": 3425,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 392.603
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12510,
          "state_bytes": 1341,
          "usage": {
            "input_tokens": 3451,
            "output_tokens": 245
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 377.952
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 370.521,
      "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": 1348.289,
      "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": 856.774
        },
        {
          "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": 353.833
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 355.572,
      "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": 1159,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1257.579,
      "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": 718.368
        },
        {
          "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": 381.816
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 429.306,
      "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": 1247,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1362.555,
      "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": 843.608
        },
        {
          "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": 355.054
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 369.145,
      "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": 1140,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1232.01,
      "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": 836.284
        },
        {
          "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": 243.355
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 360.736,
      "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": 928.491,
      "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": 851.198
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 368.675,
      "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": 762,
      "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": 838.1,
      "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": 756.072
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 368.625,
      "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": 1294,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1365.325,
      "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": 874.313
        },
        {
          "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": 362.671
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 356.006,
      "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": 1249,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1325.593,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3662,
        "output_tokens": 232
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3662,
          "output_tokens": 232
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 12472,
      "state_bytes": 1475,
      "estimated_input_usd": 0.000153804,
      "known_input_usd_lower_bound": 0.000153804,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6196,
          "state_bytes": 695,
          "usage": {
            "input_tokens": 1817,
            "output_tokens": 116
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 836.653
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6276,
          "state_bytes": 780,
          "usage": {
            "input_tokens": 1845,
            "output_tokens": 116
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 365.583
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 361.092,
      "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": 3037,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": true
      },
      "whole_operation_ms": 3110.543,
      "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": 759.515
        },
        {
          "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": 401.227
        },
        {
          "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": 360.53
        },
        {
          "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": 366.258
        },
        {
          "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": 344.998
        },
        {
          "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": 216.642
        },
        {
          "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": 355.841
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 359.86,
      "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": 0.54,
          "threshold": 0.55
        }
      ],
      "pending_label": "Scroll down in Results panel",
      "error_type": null,
      "error": null,
      "steps": 0,
      "kernel_elapsed_ms": 840,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": false
      },
      "whole_operation_ms": 915.928,
      "model_calls": 1,
      "provider_attempts": 1,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 957,
        "output_tokens": 75
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 957,
          "output_tokens": 75
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 3031,
      "state_bytes": 206,
      "estimated_input_usd": 4.0194e-05,
      "known_input_usd_lower_bound": 4.0194e-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": 836.163
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 368.999,
      "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": 0.5,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.030000000000000027,
          "threshold": 0.1
        },
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.48,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1760,
      "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": 1833.058,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9404,
        "output_tokens": 631
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9404,
          "output_tokens": 631
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32348,
      "state_bytes": 4390,
      "estimated_input_usd": 0.000394968,
      "known_input_usd_lower_bound": 0.000394968,
      "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": 855.277
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12786,
          "state_bytes": 1603,
          "usage": {
            "input_tokens": 3663,
            "output_tokens": 249
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 366.239
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13594,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 382.091
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 477.199,
      "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": 0.5,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.03999999999999998,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1629,
      "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": 1722.666,
      "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": 845.79
        },
        {
          "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": 364.978
        },
        {
          "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": 280.533
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 376.497,
      "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": 0.5,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1752,
      "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": 1848.104,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9404,
        "output_tokens": 631
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9404,
          "output_tokens": 631
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32183,
      "state_bytes": 4390,
      "estimated_input_usd": 0.000394968,
      "known_input_usd_lower_bound": 0.000394968,
      "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": 843.693
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12720,
          "state_bytes": 1603,
          "usage": {
            "input_tokens": 3663,
            "output_tokens": 249
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 383.925
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13528,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 280
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 366.441
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 364.578,
      "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": 0.39,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.06,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1605,
      "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": 1675.053,
      "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": 853.551
        },
        {
          "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": 360.671
        },
        {
          "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": 252.712
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 360.708,
      "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": 3495,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3569.56,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21172,
        "output_tokens": 1552
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21172,
          "output_tokens": 1552
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77377,
      "state_bytes": 6651,
      "estimated_input_usd": 0.000889224,
      "known_input_usd_lower_bound": 0.000889224,
      "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": 856.219
        },
        {
          "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": 363.514
        },
        {
          "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": 363.737
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12313,
          "state_bytes": 1080,
          "usage": {
            "input_tokens": 3368,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 374.13
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12366,
          "state_bytes": 1153,
          "usage": {
            "input_tokens": 3391,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 401.412
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12456,
          "state_bytes": 1232,
          "usage": {
            "input_tokens": 3425,
            "output_tokens": 244
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 389.031
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12510,
          "state_bytes": 1341,
          "usage": {
            "input_tokens": 3451,
            "output_tokens": 245
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 362.081
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 365.845,
      "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": 3114,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3189.952,
      "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": 856.114
        },
        {
          "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": 353.43
        },
        {
          "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": 410.746
        },
        {
          "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": 242.502
        },
        {
          "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": 256.22
        },
        {
          "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": 250.007
        },
        {
          "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": 397.755
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 362.354,
      "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": 1196,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1297.481,
      "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": 758.166
        },
        {
          "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": 391.94
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 379.586,
      "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": 1313,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1429.086,
      "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": 868.188
        },
        {
          "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": 393.502
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 365.182,
      "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": 1282,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1388.556,
      "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": 864.316
        },
        {
          "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": 367.266
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 375.014,
      "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": 1328,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1444.935,
      "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": 882.869
        },
        {
          "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": 391.822
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 377.575,
      "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": 841,
      "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": 916.855,
      "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": 836.47
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 411.498,
      "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": 847,
      "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": 926.131,
      "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": 841.523
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 361.146,
      "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": 1377,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1465.122,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3662,
        "output_tokens": 232
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3662,
          "output_tokens": 232
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 12472,
      "state_bytes": 1475,
      "estimated_input_usd": 0.000153804,
      "known_input_usd_lower_bound": 0.000153804,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 6196,
          "state_bytes": 695,
          "usage": {
            "input_tokens": 1817,
            "output_tokens": 116
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 910.721
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 6276,
          "state_bytes": 780,
          "usage": {
            "input_tokens": 1845,
            "output_tokens": 116
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 404.879
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "already-satisfied-controls",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 361.536,
      "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": 1143,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "confirmed": true,
        "exact_controls": true
      },
      "whole_operation_ms": 1223.68,
      "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": 857.433
        },
        {
          "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": 230.458
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 362.812,
      "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": 1237,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": false
      },
      "whole_operation_ms": 1314.216,
      "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": 837.432
        },
        {
          "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": 356.176
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "nested-scroll-target",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 359.482,
      "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": 3379,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "target_effect": true
      },
      "whole_operation_ms": 3455.259,
      "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": 734.28
        },
        {
          "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": 464.33
        },
        {
          "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": 428.629
        },
        {
          "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": 345.985
        },
        {
          "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": 363.882
        },
        {
          "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": 358.567
        },
        {
          "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": 428.767
        }
      ],
      "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."
  ]
}
