{
  "schema_version": 1,
  "kind": "live-browser-pinned-ab",
  "created_at": "2026-10-04T07:33:50.168943+00:00",
  "commits": {
    "baseline": "39d0bdc21dd3b17de452f413181f422a4c84af18",
    "treatment": "working-tree-snapshot"
  },
  "source_sha256": {
    "baseline": "329682d8b5e60d9ea81fe6a552ddc11af952c65f4b322db6f9bf42db77a3e18f",
    "treatment": "4a73f6d212df8f71a41348902558015627935bfd2f972e240ff2f69be8d449d0"
  },
  "repeats": 3,
  "tasks": [
    "search-filter",
    "buy-pauses-before-order",
    "contact-form-scroll",
    "shadow-dom",
    "same-origin-frame",
    "delete-is-gated"
  ],
  "fixture_sha256": "2962326329e311f38b3cbe9d74881aedd436db33f23b6f74c1dfb166cd275fa5",
  "runtime": {
    "python": "3.14.2",
    "system": "Darwin"
  },
  "whole_benchmark_elapsed_ms": 95992.813,
  "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": 18,
      "passed": 11,
      "correct_completions": 8,
      "correct_gates": 3,
      "review_or_failure": 7,
      "statuses": {
        "needs_review": 7,
        "done": 8,
        "needs_confirmation": 3
      },
      "false_completions": 0,
      "wrong_actions": 0,
      "model_calls": 56,
      "provider_attempts": 56,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 154865,
        "output_tokens": 10824
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 154865,
          "output_tokens": 10824
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "whole_operation_total_ms": 33134.114,
      "whole_operation_median_ms": 1495.221,
      "request_bytes": 542915,
      "state_bytes": 59499,
      "estimated_input_usd": 0.00650433
    },
    "treatment": {
      "runs": 18,
      "passed": 9,
      "correct_completions": 6,
      "correct_gates": 3,
      "review_or_failure": 9,
      "statuses": {
        "needs_review": 9,
        "done": 6,
        "needs_confirmation": 3
      },
      "false_completions": 0,
      "wrong_actions": 0,
      "model_calls": 48,
      "provider_attempts": 48,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 124125,
        "output_tokens": 8610
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 124125,
          "output_tokens": 8610
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "whole_operation_total_ms": 29426.042,
      "whole_operation_median_ms": 1497.698,
      "request_bytes": 436932,
      "state_bytes": 46446,
      "estimated_input_usd": 0.00521325
    }
  },
  "runs": [
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 843.269,
      "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
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.040000000000000036,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 1947,
      "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": 2017.912,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 13630,
        "output_tokens": 922
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 13630,
          "output_tokens": 922
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 46283,
      "state_bytes": 6564,
      "estimated_input_usd": 0.00057246,
      "known_input_usd_lower_bound": 0.00057246,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5968,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 865.22
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12903,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 390.558
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13711,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 238.402
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13701,
          "state_bytes": 2156,
          "usage": {
            "input_tokens": 4120,
            "output_tokens": 278
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 233.501
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 373.523,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "operation",
          "reason": "below_min_top_probability",
          "actual": 0.52,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.06,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category → Shoes",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1717,
      "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": 1789.054,
      "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": 762.714
        },
        {
          "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": 400.292
        },
        {
          "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": 401.953
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 370.671,
      "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": 0.43,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.07,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 2101,
      "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": 2174.831,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32417,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5935,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 967.028
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12837,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 483.02
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13645,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 504.142
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 427.802,
      "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": 0.47,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1609,
      "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": 1682.961,
      "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": 852.981
        },
        {
          "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": 358.967
        },
        {
          "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": 249.071
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 360.188,
      "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": 3282,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3358.11,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21378,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21378,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77772,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000897876,
      "known_input_usd_lower_bound": 0.000897876,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 9038,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2429,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 914.478
        },
        {
          "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": 230.522
        },
        {
          "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": 228.613
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12417,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3421,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 402.487
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12463,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3442,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 365.26
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12553,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3476,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 368.12
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12607,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3502,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 397.03
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 453.453,
      "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": "operation",
          "reason": "below_min_top_probability",
          "actual": 0.51,
          "threshold": 0.55
        },
        {
          "question": "operation",
          "reason": "below_min_margin",
          "actual": 0.06,
          "threshold": 0.1
        }
      ],
      "pending_label": "I agree to be contacted *",
      "error_type": null,
      "error": null,
      "steps": 4,
      "kernel_elapsed_ms": 2624,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": false,
        "exact_form_values": false
      },
      "whole_operation_ms": 2698.657,
      "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": 848.208
        },
        {
          "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": 480.655
        },
        {
          "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": 232.33
        },
        {
          "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": 445.646
        },
        {
          "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": 362.225
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 361.089,
      "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": 1171,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1272.985,
      "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": 726.762
        },
        {
          "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": 396.304
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 362.542,
      "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": 1283,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1381.422,
      "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": 872.144
        },
        {
          "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": 361.852
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 355.67,
      "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": 1183,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1278.208,
      "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": 857.433
        },
        {
          "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": 267.481
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 359.107,
      "passed": true,
      "correct_completion": true,
      "correct_gate": false,
      "review_or_failure": false,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "done",
      "reason": null,
      "review_reasons": [],
      "pending_label": null,
      "error_type": null,
      "error": null,
      "steps": 1,
      "kernel_elapsed_ms": 1290,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1400.81,
      "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": 854.975
        },
        {
          "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": 376.808
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 384.671,
      "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": 883,
      "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": 973.267,
      "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": 878.108
        }
      ],
      "arm": "baseline",
      "repeat": 1
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 391.829,
      "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": 745,
      "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": 819.046,
      "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": 739.784
        }
      ],
      "arm": "treatment",
      "repeat": 1
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 369.304,
      "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": 0.48,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1636,
      "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": 1712.346,
      "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": 863.962
        },
        {
          "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": 394.906
        },
        {
          "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": 234.733
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 379.704,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.45,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.040000000000000036,
          "threshold": 0.1
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 2010,
      "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": 2091.215,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 13630,
        "output_tokens": 922
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 13630,
          "output_tokens": 922
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 46283,
      "state_bytes": 6564,
      "estimated_input_usd": 0.00057246,
      "known_input_usd_lower_bound": 0.00057246,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5968,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 870.485
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12903,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 433.292
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13711,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 259.744
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13701,
          "state_bytes": 2156,
          "usage": {
            "input_tokens": 4120,
            "output_tokens": 278
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 242.003
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 361.704,
      "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": 0.47000000000000003,
          "threshold": 0.55
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1697,
      "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": 1768.032,
      "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": 891.097
        },
        {
          "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": 377.803
        },
        {
          "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": 294.189
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 378.308,
      "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": 0.42,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.03999999999999998,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1523,
      "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": 1593.252,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32417,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5935,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 747.984
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12837,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 367.776
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13645,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 259.883
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 381.891,
      "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": 0.02999999999999997,
          "threshold": 0.1
        }
      ],
      "pending_label": "I agree to be contacted *",
      "error_type": null,
      "error": null,
      "steps": 4,
      "kernel_elapsed_ms": 2360,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": false,
        "exact_form_values": false
      },
      "whole_operation_ms": 2442.723,
      "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": 850.68
        },
        {
          "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": 241.391
        },
        {
          "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": 261.797
        },
        {
          "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": 379.237
        },
        {
          "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": 365.87
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 362.137,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.52,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.040000000000000036,
          "threshold": 0.1
        }
      ],
      "pending_label": "Save draft",
      "error_type": null,
      "error": null,
      "steps": 5,
      "kernel_elapsed_ms": 2958,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": false,
        "exact_form_values": true
      },
      "whole_operation_ms": 3034.746,
      "model_calls": 6,
      "provider_attempts": 6,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 17876,
        "output_tokens": 1328
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 17876,
          "output_tokens": 1328
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 65165,
      "state_bytes": 5337,
      "estimated_input_usd": 0.000750792,
      "known_input_usd_lower_bound": 0.000750792,
      "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": 847.881
        },
        {
          "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": 367.558
        },
        {
          "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": 298.573
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12417,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3421,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 380.457
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12463,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3442,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 406.679
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12553,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3476,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 358.342
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 368.366,
      "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": 1246,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1358.48,
      "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": 843.677
        },
        {
          "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": 351.924
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 360.435,
      "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": 1298,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1397.19,
      "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": 854.769
        },
        {
          "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": 389.876
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 367.6,
      "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": 1275,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1382.41,
      "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": 857.502
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 5788,
          "state_bytes": 544,
          "usage": {
            "input_tokens": 1661,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 362.063
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 474.414,
      "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": 1262,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1361.432,
      "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": 855.912
        },
        {
          "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": 361.002
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 525.829,
      "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": 906,
      "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": 1091.887,
      "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": 901.076
        }
      ],
      "arm": "treatment",
      "repeat": 2
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 686.106,
      "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": 1163,
      "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": 1311.895,
      "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": 1158.236
        }
      ],
      "arm": "baseline",
      "repeat": 2
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 873.613,
      "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
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.040000000000000036,
          "threshold": 0.1
        }
      ],
      "pending_label": "Category",
      "error_type": null,
      "error": null,
      "steps": 3,
      "kernel_elapsed_ms": 2293,
      "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": 2374.393,
      "model_calls": 4,
      "provider_attempts": 4,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 13630,
        "output_tokens": 922
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 13630,
          "output_tokens": 922
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 46283,
      "state_bytes": 6564,
      "estimated_input_usd": 0.00057246,
      "known_input_usd_lower_bound": 0.00057246,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5968,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 1168.616
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12903,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 389.575
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13711,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 286
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 305.145
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13701,
          "state_bytes": 2156,
          "usage": {
            "input_tokens": 4120,
            "output_tokens": 278
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 257.705
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "search-filter",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 576.18,
      "passed": false,
      "correct_completion": false,
      "correct_gate": false,
      "review_or_failure": true,
      "false_completion": false,
      "wrong_actions": 0,
      "status": "needs_review",
      "reason": "uncertain_decision",
      "review_reasons": [
        {
          "question": "click_target",
          "reason": "below_min_top_probability",
          "actual": 0.45,
          "threshold": 0.55
        },
        {
          "question": "click_target",
          "reason": "below_min_margin",
          "actual": 0.06,
          "threshold": 0.1
        }
      ],
      "pending_label": "red shoes",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1920,
      "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": 2115.784,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9404,
        "output_tokens": 630
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9404,
          "output_tokens": 630
        }
      },
      "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": 1102.32
        },
        {
          "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": 461.181
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13594,
          "state_bytes": 2084,
          "usage": {
            "input_tokens": 4064,
            "output_tokens": 279
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 234.046
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 593.794,
      "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": 0.010000000000000009,
          "threshold": 0.1
        }
      ],
      "pending_label": "In stock only",
      "error_type": null,
      "error": null,
      "steps": 2,
      "kernel_elapsed_ms": 1975,
      "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": 2048.003,
      "model_calls": 3,
      "provider_attempts": 3,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 9510,
        "output_tokens": 645
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 9510,
          "output_tokens": 645
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 32417,
      "state_bytes": 4408,
      "estimated_input_usd": 0.00039942,
      "known_input_usd_lower_bound": 0.00039942,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5935,
          "state_bytes": 703,
          "usage": {
            "input_tokens": 1677,
            "output_tokens": 102
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 1055.667
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12837,
          "state_bytes": 1612,
          "usage": {
            "input_tokens": 3716,
            "output_tokens": 256
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 368.125
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 13645,
          "state_bytes": 2093,
          "usage": {
            "input_tokens": 4117,
            "output_tokens": 287
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 408.017
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "buy-pauses-before-order",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 378.496,
      "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": 0.42,
          "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": 1522,
      "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": 1594.587,
      "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": 762.537
        },
        {
          "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": 374.424
        },
        {
          "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": 241.349
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 361.8,
      "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": 3276,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": true,
        "exact_form_values": true
      },
      "whole_operation_ms": 3349.189,
      "model_calls": 7,
      "provider_attempts": 7,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 21378,
        "output_tokens": 1580
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 21378,
          "output_tokens": 1580
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 77772,
      "state_bytes": 6687,
      "estimated_input_usd": 0.000897876,
      "known_input_usd_lower_bound": 0.000897876,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 9038,
          "state_bytes": 505,
          "usage": {
            "input_tokens": 2429,
            "output_tokens": 186
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 832.958
        },
        {
          "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": 240.761
        },
        {
          "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": 359.411
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12417,
          "state_bytes": 1089,
          "usage": {
            "input_tokens": 3421,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 411.947
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12463,
          "state_bytes": 1162,
          "usage": {
            "input_tokens": 3442,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 371.249
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12553,
          "state_bytes": 1241,
          "usage": {
            "input_tokens": 3476,
            "output_tokens": 251
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 390.752
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12607,
          "state_bytes": 1350,
          "usage": {
            "input_tokens": 3502,
            "output_tokens": 252
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 361.729
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "contact-form-scroll",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 416.829,
      "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": 0.06,
          "threshold": 0.1
        }
      ],
      "pending_label": "Message",
      "error_type": null,
      "error": null,
      "steps": 4,
      "kernel_elapsed_ms": 2526,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "saved_draft": false,
        "exact_form_values": false
      },
      "whole_operation_ms": 2614.145,
      "model_calls": 5,
      "provider_attempts": 5,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 14296,
        "output_tokens": 1064
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 14296,
          "output_tokens": 1064
        }
      },
      "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": 837.112
        },
        {
          "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": 349.537
        },
        {
          "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": 362.263
        },
        {
          "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": 358.197
        },
        {
          "model": "jev-1.13.0",
          "request_bytes": 12366,
          "state_bytes": 1153,
          "usage": {
            "input_tokens": 3391,
            "output_tokens": 245
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 372.099
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 401.603,
      "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": 1200,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1314.331,
      "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": 862.534
        },
        {
          "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": 292.33
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "shadow-dom",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 391.873,
      "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": 1171,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1287.44,
      "model_calls": 2,
      "provider_attempts": 2,
      "provider_retries": 0,
      "failed_provider_calls": 0,
      "known_usage": {
        "input_tokens": 3268,
        "output_tokens": 218
      },
      "model_usage": {
        "jev-1.13.0": {
          "input_tokens": 3268,
          "output_tokens": 218
        }
      },
      "usage_complete": true,
      "unknown_usage_attempts": 0,
      "request_bytes": 11352,
      "state_bytes": 1015,
      "estimated_input_usd": 0.000137256,
      "known_input_usd_lower_bound": 0.000137256,
      "calls": [
        {
          "model": "jev-1.13.0",
          "request_bytes": 5642,
          "state_bytes": 471,
          "usage": {
            "input_tokens": 1621,
            "output_tokens": 109
          },
          "usage_complete": true,
          "unknown_usage_attempts": 0,
          "attempts": 1,
          "elapsed_ms": 846.966
        },
        {
          "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": 275.422
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 426.545,
      "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": 1098,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1199.516,
      "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": 810.606
        },
        {
          "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": 235.737
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "same-origin-frame",
      "runtime_version": "0.4.0",
      "expected": "done",
      "browser_startup_ms": 430.753,
      "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": 1192,
      "kernel_usage_complete": true,
      "kernel_unknown_usage_attempts": 0,
      "independent_checks": {
        "expected_widget_effect": true
      },
      "whole_operation_ms": 1300.087,
      "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": 848.881
        },
        {
          "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": 284.57
        }
      ],
      "arm": "treatment",
      "repeat": 3
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 744.314,
      "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": 897,
      "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": 983.639,
      "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": 890.847
        }
      ],
      "arm": "baseline",
      "repeat": 3
    },
    {
      "task": "delete-is-gated",
      "runtime_version": "0.4.0",
      "expected": "needs_confirmation",
      "browser_startup_ms": 848.57,
      "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": 894,
      "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": 986.171,
      "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": 888.198
        }
      ],
      "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."
  ]
}
