{
  "mode": "offline-synthetic-planning-diagnostics",
  "baseline_commit": "bb808e1472d3894e276689115574f126c193c52a",
  "candidate": "working tree at invocation; published with its source commit",
  "baseline_source_sha256": {
    "src/apixly_jev_harness/__init__.py": "10df257ec1770fce69e97a0b5199d9b034092f5557f6a6a1fb2aa61b42e7efe2",
    "src/apixly_jev_harness/backend.py": "e4b4c80d88c63648c1392f41eb8796430fcbce6f81d1061c71519d0759432eda",
    "src/apixly_jev_harness/callbacks.py": "2d27ad4540d910354c54de7566546b4c9d99b5a5fa87705a1905885349078497",
    "src/apixly_jev_harness/cli.py": "8d9434bf8a8b539163194a84e3f099761d08ef32201f99cbce46c139f3726d94",
    "src/apixly_jev_harness/context.py": "b9d76abb33f8a1ebe8731fe6d5fe6811445132a6046448fc0e2cc871507d50ea",
    "src/apixly_jev_harness/diagnostics.py": "24950b4d834d9e6c902996c7bcddfa208944e42300ae062e9ebc2a44f75a091f",
    "src/apixly_jev_harness/engine.py": "19d809ee7af67586b03a472af49908f6ec278231285d16db161a540b3f0247a6",
    "src/apixly_jev_harness/output.py": "94be8b482e4279e29aca128a35caf54744dbe8db8188f14905488f8ddde93a0b",
    "src/apixly_jev_harness/surfaces.py": "65ce73edfe9f6a3c16de373b3a1b85671c3d2cc661f3a50dd906a15721c7d8f1",
    "src/apixly_jev_harness/task.py": "0f5801e9fdb5626e1db65d269c715e8bcf692ebaa3a91f46b9945158b4bbb6db",
    "src/apixly_jev_harness/task_cli.py": "bc2aeea2af51fe26efc4528efc9f7849078030a1cea1de2420ff1e98dcabf951",
    "src/apixly_jev_harness/verify.py": "97d2a178902291094f4e16db08732748d234b8c5ad8dd49a55607c8d195f1ed2"
  },
  "candidate_source_sha256": {
    "src/apixly_jev_harness/__init__.py": "79a4ea23ea1469f218c5d6f4ee8567f0cf12f77dbd658f4afb57daa90c2a3039",
    "src/apixly_jev_harness/backend.py": "0ddfab6b270d8219a66614630e70aa10a74cb531f8075d07c255df145725ed5f",
    "src/apixly_jev_harness/callbacks.py": "2d27ad4540d910354c54de7566546b4c9d99b5a5fa87705a1905885349078497",
    "src/apixly_jev_harness/cli.py": "8d9434bf8a8b539163194a84e3f099761d08ef32201f99cbce46c139f3726d94",
    "src/apixly_jev_harness/context.py": "b9d76abb33f8a1ebe8731fe6d5fe6811445132a6046448fc0e2cc871507d50ea",
    "src/apixly_jev_harness/diagnostics.py": "9e8e98dc4b103c1a6687dbd31bd2dd01739309c7e4021b9ecb0f28b6976a0f71",
    "src/apixly_jev_harness/engine.py": "19d809ee7af67586b03a472af49908f6ec278231285d16db161a540b3f0247a6",
    "src/apixly_jev_harness/output.py": "94be8b482e4279e29aca128a35caf54744dbe8db8188f14905488f8ddde93a0b",
    "src/apixly_jev_harness/surfaces.py": "65ce73edfe9f6a3c16de373b3a1b85671c3d2cc661f3a50dd906a15721c7d8f1",
    "src/apixly_jev_harness/task.py": "0f5801e9fdb5626e1db65d269c715e8bcf692ebaa3a91f46b9945158b4bbb6db",
    "src/apixly_jev_harness/task_cli.py": "9cf30e5ed50680ac3b0212bdd8118d67f4272cc7690941a6e57e48fd2ccf850d",
    "src/apixly_jev_harness/verify.py": "97d2a178902291094f4e16db08732748d234b8c5ad8dd49a55607c8d195f1ed2"
  },
  "runner_sha256": "e21758013f7cb1ece4a5b8b871dd4b5c8e7ee8e9c5a3f9ab893e68901cee18cc",
  "python_version": "3.14.2",
  "cases": {
    "direct_ascii_60k": {
      "mode": "direct",
      "character": "A",
      "text_repetitions": 60000,
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "direct_chinese_60k": {
      "mode": "direct",
      "character": "\u4e2d",
      "text_repetitions": 60000,
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "direct_254_choices": {
      "mode": "direct",
      "records": 254,
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "direct_missing_record_field": {
      "mode": "direct",
      "required_record_fields": [
        "source.required"
      ],
      "plan_code": "request_needs_context",
      "typed_status": "NEEDS_CONTEXT"
    },
    "direct_small_mock_choice": {
      "mode": "direct",
      "mock_response": "choice"
    },
    "loop_tool_oversize": {
      "mode": "loop",
      "location": "tool",
      "text_repetitions": 60000,
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "loop_parameter_oversize": {
      "mode": "loop",
      "location": "parameter",
      "text_repetitions": 60000,
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "prior_unknown_then_parameter": {
      "mode": "loop",
      "location": "parameter",
      "text_repetitions": 60000,
      "mock_response": "retry_then_choice",
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "fallback_then_effect_error": {
      "mode": "loop",
      "location": "fallback",
      "text_repetitions": 60000,
      "plan_code": "request_needs_narrowing",
      "typed_status": "NEEDS_NARROWING"
    },
    "unknown_plan_deferred": {
      "mode": "direct",
      "plan_stub": "unknown",
      "plan_code": "request_planning_review",
      "typed_status": "UNAVAILABLE"
    },
    "unsafe_plan_enum_injection": {
      "mode": "direct",
      "plan_stub": "injected",
      "plan_code": "request_needs_context",
      "typed_status": "NEEDS_CONTEXT"
    },
    "generic_spec_value_error": {
      "mode": "direct",
      "context_repetitions": 17000,
      "expected_exception": "ValueError"
    },
    "mock_http_413_loop": {
      "mode": "loop",
      "mock_response": "http_413"
    },
    "mock_http_413_direct": {
      "mode": "direct",
      "mock_response": "http_413"
    }
  },
  "behavior_parity": true,
  "reproduction": "python benchmarks/planning_diagnostics.py --output local-results/planning-diagnostics-ab.json",
  "scope": "Fourteen fixed synthetic fixtures, three paired repeats with alternating arm order. Actual pinned core.plan handles oversized ASCII/Chinese records, 254 options, required-record fields and dependent parameter descriptions. Small/HTTP/retry fixtures execute the pinned typed batch and provider Client only through explicit httpx.MockTransport with a synthetic key; real HTTPTransport is forbidden. Mock dispatches and HTTP attempts are separate from actual model usage, which is zero if the guard is never hit. Two injected plan fixtures test safe enum projection; a generic 17,000-character spec ValueError remains an exception. Unknown usage after a mocked retry or HTTP 413 remains unknown in reported mock telemetry.",
  "negative_results": "New fixed planning metadata and reason codes add returned bytes and local projection work. Request fit limits, option counts, prompts and loop context limits are unchanged; typed NEEDS_CONTEXT is preserved instead of being folded into NEEDS_NARROWING, and unrecognized statuses become UNAVAILABLE. End-state/selection/effect parity excludes that deliberate typed-status correction and diagnostic fields. No semantic quality, speed, real-provider fit, live HTTP behavior, fee benefit or automatic retry improvement is claimed. The only retry is the pinned SDK's existing bounded policy exercised by synthetic MockTransport. A trusted fallback clears the current failure diagnostic but keeps its historical call metadata; later effects remain execution_uncertain.",
  "cost_conversion": "No tariff or savings assumed. Actual model usage is zero only with no forbidden real transport attempts; a guard hit records unknown actual usage and fails the fixture. Reported mock usage and unknown mock attempts never represent paid usage. Real whole-operation timing includes synthetic registration, actual planning, mocked typed batch/provider validation, bounded mock transport attempts, permitted effect and private journal serialization; excludes process/import/client-fixture setup and terminal-analysis/source hashing. Returned context bytes exclude local archive paths; raw candidate bodies, injected detail and response text are omitted.",
  "summary": {
    "baseline": {
      "runs": 42,
      "correct": 12,
      "behavior_correct": 42,
      "planning_correct": 12,
      "diagnostic_correct": 33,
      "typed_status_correct": 33,
      "mock_batch_dispatches": 18,
      "mock_http_attempts": 21,
      "effects": 3,
      "intent_events": 3,
      "real_network_attempts": 0,
      "raw_marker_exposures": 0,
      "median_whole_operation_seconds": 0.002040020495769568,
      "returned_context_bytes": 33295,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      }
    },
    "candidate": {
      "runs": 42,
      "correct": 42,
      "behavior_correct": 42,
      "planning_correct": 42,
      "diagnostic_correct": 42,
      "typed_status_correct": 42,
      "mock_batch_dispatches": 18,
      "mock_http_attempts": 21,
      "effects": 3,
      "intent_events": 3,
      "real_network_attempts": 0,
      "raw_marker_exposures": 0,
      "median_whole_operation_seconds": 0.0020040209928993136,
      "returned_context_bytes": 37777,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      }
    }
  },
  "rows": [
    {
      "case": "direct_ascii_60k",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "A",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0015078750002430752,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "direct_ascii_60k",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "A",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002654708005138673,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "direct_ascii_60k",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "A",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0011460420064395294,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "direct_ascii_60k",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "A",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0026622919976944104,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "direct_ascii_60k",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "A",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0011713330022757873,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "direct_ascii_60k",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "A",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0022605410049436614,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "direct_chinese_60k",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "\u4e2d",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0017705000063870102,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "direct_chinese_60k",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "\u4e2d",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001354541993350722,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "direct_chinese_60k",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "\u4e2d",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.00235574999533128,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "direct_chinese_60k",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "\u4e2d",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0013082079967716709,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "direct_chinese_60k",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "\u4e2d",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0026687500067055225,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "direct_chinese_60k",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "character": "\u4e2d",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0020736669976031408,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "direct_254_choices",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "records": 254,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 254,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "source_count": 254
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.00219462500535883,
      "returned_context_bytes": 1896,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "direct_254_choices",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "records": 254,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 254,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "source_count": 254
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.003182542001013644,
      "returned_context_bytes": 2045,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "direct_254_choices",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "records": 254,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 254,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "source_count": 254
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0017458329966757447,
      "returned_context_bytes": 2045,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "direct_254_choices",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "records": 254,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 254,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "source_count": 254
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.002716457995120436,
      "returned_context_bytes": 1896,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "direct_254_choices",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "records": 254,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 254,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "source_count": 254
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0016605419950792566,
      "returned_context_bytes": 1896,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "direct_254_choices",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "records": 254,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 254,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 254
            },
            "source_count": 254
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0017830839933594689,
      "returned_context_bytes": 2045,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "direct_missing_record_field",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "required_record_fields": [
          "source.required"
        ],
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0010029999975813553,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "direct_missing_record_field",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "required_record_fields": [
          "source.required"
        ],
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_context"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001676375002716668,
      "returned_context_bytes": 385,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "direct_missing_record_field",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "required_record_fields": [
          "source.required"
        ],
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_context"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0011157499975524843,
      "returned_context_bytes": 385,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "direct_missing_record_field",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "required_record_fields": [
          "source.required"
        ],
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0029814580047968775,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "direct_missing_record_field",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "required_record_fields": [
          "source.required"
        ],
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0010848749952856451,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "direct_missing_record_field",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "required_record_fields": [
          "source.required"
        ],
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_context"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0008291249978356063,
      "returned_context_bytes": 385,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "direct_small_mock_choice",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "choice"
      },
      "expected": {
        "status": "admitted",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "admitted",
        "exception_type": null,
        "complete": true,
        "selected_count": 1,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "OK": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0020724579953821376,
      "returned_context_bytes": 835,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "direct_small_mock_choice",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "choice"
      },
      "expected": {
        "status": "admitted",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "admitted",
        "exception_type": null,
        "complete": true,
        "selected_count": 1,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "OK": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0012732920004054904,
      "returned_context_bytes": 835,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "direct_small_mock_choice",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "choice"
      },
      "expected": {
        "status": "admitted",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "admitted",
        "exception_type": null,
        "complete": true,
        "selected_count": 1,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "OK": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001301915996009484,
      "returned_context_bytes": 835,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "direct_small_mock_choice",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "choice"
      },
      "expected": {
        "status": "admitted",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "admitted",
        "exception_type": null,
        "complete": true,
        "selected_count": 1,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "OK": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002146958009689115,
      "returned_context_bytes": 835,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "direct_small_mock_choice",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "choice"
      },
      "expected": {
        "status": "admitted",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "admitted",
        "exception_type": null,
        "complete": true,
        "selected_count": 1,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "OK": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001578458002768457,
      "returned_context_bytes": 835,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "direct_small_mock_choice",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "choice"
      },
      "expected": {
        "status": "admitted",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "admitted",
        "exception_type": null,
        "complete": true,
        "selected_count": 1,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "OK": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0015202920039882883,
      "returned_context_bytes": 835,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "loop_tool_oversize",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "tool",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "source_count": 4
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.003307292005047202,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "loop_tool_oversize",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "tool",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0031136670004343614,
      "returned_context_bytes": 841,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "loop_tool_oversize",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "tool",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.003281541998148896,
      "returned_context_bytes": 839,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "loop_tool_oversize",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "tool",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "source_count": 4
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.002007582996156998,
      "returned_context_bytes": 688,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "loop_tool_oversize",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "tool",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "source_count": 4
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0027126249915454537,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "loop_tool_oversize",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "tool",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 4
            },
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0027865419979207218,
      "returned_context_bytes": 839,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "loop_parameter_oversize",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2687,
            "request_json_utf8_bytes": 2687,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.002736292008194141,
      "returned_context_bytes": 1626,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "loop_parameter_oversize",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.006056749989511445,
      "returned_context_bytes": 1776,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "loop_parameter_oversize",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0033659999899100512,
      "returned_context_bytes": 1777,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "loop_parameter_oversize",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0025123749946942553,
      "returned_context_bytes": 1624,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "loop_parameter_oversize",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0026291670073987916,
      "returned_context_bytes": 1626,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "loop_parameter_oversize",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0026914999907603487,
      "returned_context_bytes": 1777,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "prior_unknown_then_parameter",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "mock_response": "retry_then_choice",
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "usage_complete": false,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "mock_request_sizes": [
          {
            "request_json_characters": 2686,
            "request_json_utf8_bytes": 2686,
            "question_count": 1,
            "criterion_count": 6
          },
          {
            "request_json_characters": 2686,
            "request_json_utf8_bytes": 2686,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0037216250057099387,
      "returned_context_bytes": 1548,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "prior_unknown_then_parameter",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "mock_response": "retry_then_choice",
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "usage_complete": false,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          },
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002934958989499137,
      "returned_context_bytes": 1703,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "prior_unknown_then_parameter",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "mock_response": "retry_then_choice",
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "usage_complete": false,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          },
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0043727919983211905,
      "returned_context_bytes": 1702,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "prior_unknown_then_parameter",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "mock_response": "retry_then_choice",
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "usage_complete": false,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          },
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0033062500006053597,
      "returned_context_bytes": 1551,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "prior_unknown_then_parameter",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "mock_response": "retry_then_choice",
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "usage_complete": false,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          },
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "matches_expected": false
        },
        "diagnostic_correct": false,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0027609170065261424,
      "returned_context_bytes": 1549,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "prior_unknown_then_parameter",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "parameter",
        "text_repetitions": 60000,
        "mock_response": "retry_then_choice",
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "usage_complete": false,
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 2,
        "mock_request_sizes": [
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          },
          {
            "request_json_characters": 2688,
            "request_json_utf8_bytes": 2688,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": {
          "stage": "parameter_selection",
          "reason_codes": [
            "request_needs_narrowing"
          ],
          "automatic_retry": false,
          "tool_id": "inspect",
          "parameter": "target"
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002880291998735629,
      "returned_context_bytes": 1703,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "fallback_then_effect_error",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "fallback",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "execution_uncertain",
        "effects": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "execution_uncertain",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 1,
        "intent_events": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2125,
            "request_json_utf8_bytes": 2125,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0030075410031713545,
      "returned_context_bytes": 1485,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "fallback_then_effect_error",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "fallback",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "execution_uncertain",
        "effects": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "execution_uncertain",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 1,
        "intent_events": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2126,
            "request_json_utf8_bytes": 2126,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002689832996111363,
      "returned_context_bytes": 1633,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "fallback_then_effect_error",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "fallback",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "execution_uncertain",
        "effects": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "execution_uncertain",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 1,
        "intent_events": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2126,
            "request_json_utf8_bytes": 2126,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002639875005115755,
      "returned_context_bytes": 1633,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "fallback_then_effect_error",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "fallback",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "execution_uncertain",
        "effects": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "execution_uncertain",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 1,
        "intent_events": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2126,
            "request_json_utf8_bytes": 2126,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.002743124990956858,
      "returned_context_bytes": 1485,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "fallback_then_effect_error",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "fallback",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "execution_uncertain",
        "effects": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "execution_uncertain",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 1,
        "intent_events": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2125,
            "request_json_utf8_bytes": 2125,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.004348458998720162,
      "returned_context_bytes": 1485,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "fallback_then_effect_error",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "location": "fallback",
        "text_repetitions": 60000,
        "plan_code": "request_needs_narrowing",
        "typed_status": "NEEDS_NARROWING"
      },
      "expected": {
        "status": "execution_uncertain",
        "effects": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "execution_uncertain",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 0,
        "effects": 1,
        "intent_events": 1,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2126,
            "request_json_utf8_bytes": 2126,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "OK": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          },
          {
            "stage": "parameter_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_narrowing"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          },
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 5,
          "output_tokens": 1
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.003129542004899122,
      "returned_context_bytes": 1634,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "unknown_plan_deferred",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "unknown",
        "plan_code": "request_planning_review",
        "typed_status": "UNAVAILABLE"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "UNRECOGNIZED": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0007546659908257425,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "unknown_plan_deferred",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "unknown",
        "plan_code": "request_planning_review",
        "typed_status": "UNAVAILABLE"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_planning_review"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "UNRECOGNIZED": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0011167080083396286,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "unknown_plan_deferred",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "unknown",
        "plan_code": "request_planning_review",
        "typed_status": "UNAVAILABLE"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_planning_review"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "UNRECOGNIZED": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0007550000009359792,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "unknown_plan_deferred",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "unknown",
        "plan_code": "request_planning_review",
        "typed_status": "UNAVAILABLE"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "UNRECOGNIZED": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0008060410036705434,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "unknown_plan_deferred",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "unknown",
        "plan_code": "request_planning_review",
        "typed_status": "UNAVAILABLE"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "UNRECOGNIZED": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0007622920093126595,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "unknown_plan_deferred",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "unknown",
        "plan_code": "request_planning_review",
        "typed_status": "UNAVAILABLE"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_planning_review"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "UNRECOGNIZED": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0010197499941568822,
      "returned_context_bytes": 387,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "unsafe_plan_enum_injection",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "injected",
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.001181624989840202,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "unsafe_plan_enum_injection",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "injected",
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_context"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001775249998900108,
      "returned_context_bytes": 385,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "unsafe_plan_enum_injection",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "injected",
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_context"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001266374994884245,
      "returned_context_bytes": 385,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "unsafe_plan_enum_injection",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "injected",
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0007857910095481202,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "unsafe_plan_enum_injection",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "injected",
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_NARROWING": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": false
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": false,
        "typed_status_correct": false,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": false,
      "whole_operation_seconds": 0.0007378330046776682,
      "returned_context_bytes": 238,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "unsafe_plan_enum_injection",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "plan_stub": "injected",
        "plan_code": "request_needs_context",
        "typed_status": "NEEDS_CONTEXT"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": true,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "planning": {
              "scope": "current_selection",
              "phase": "request_planning",
              "inference_dispatched": false,
              "reason_codes": [
                "request_needs_context"
              ]
            },
            "planning_raw_marker_exposures": 0,
            "requests_reported": 0,
            "usage_complete": true,
            "usage_known": true,
            "provider_failure_codes": [],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 0,
            "deferred_status_counts": {
              "NEEDS_CONTEXT": 2
            },
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": true,
        "usage_known": true,
        "reported_mock_usage": {
          "input_tokens": 0,
          "output_tokens": 0
        },
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0012437080004019663,
      "returned_context_bytes": 385,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "generic_spec_value_error",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "context_repetitions": 17000,
        "expected_exception": "ValueError"
      },
      "expected": {
        "status": "raised_value_error",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": null,
        "diagnostic": null
      },
      "result": {
        "status": "raised_value_error",
        "exception_type": "ValueError",
        "complete": null,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [],
        "plans": [],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": null,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0010591669997666031,
      "returned_context_bytes": 32,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "generic_spec_value_error",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "context_repetitions": 17000,
        "expected_exception": "ValueError"
      },
      "expected": {
        "status": "raised_value_error",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": null,
        "diagnostic": null
      },
      "result": {
        "status": "raised_value_error",
        "exception_type": "ValueError",
        "complete": null,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [],
        "plans": [],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": null,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0013340830046217889,
      "returned_context_bytes": 32,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "generic_spec_value_error",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "context_repetitions": 17000,
        "expected_exception": "ValueError"
      },
      "expected": {
        "status": "raised_value_error",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": null,
        "diagnostic": null
      },
      "result": {
        "status": "raised_value_error",
        "exception_type": "ValueError",
        "complete": null,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [],
        "plans": [],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": null,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0008446249994449317,
      "returned_context_bytes": 32,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "generic_spec_value_error",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "context_repetitions": 17000,
        "expected_exception": "ValueError"
      },
      "expected": {
        "status": "raised_value_error",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": null,
        "diagnostic": null
      },
      "result": {
        "status": "raised_value_error",
        "exception_type": "ValueError",
        "complete": null,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [],
        "plans": [],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": null,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0012756659998558462,
      "returned_context_bytes": 32,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "generic_spec_value_error",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "context_repetitions": 17000,
        "expected_exception": "ValueError"
      },
      "expected": {
        "status": "raised_value_error",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": null,
        "diagnostic": null
      },
      "result": {
        "status": "raised_value_error",
        "exception_type": "ValueError",
        "complete": null,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [],
        "plans": [],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": null,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0007989169971551746,
      "returned_context_bytes": 32,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "generic_spec_value_error",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "context_repetitions": 17000,
        "expected_exception": "ValueError"
      },
      "expected": {
        "status": "raised_value_error",
        "effects": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "usage_complete": null,
        "diagnostic": null
      },
      "result": {
        "status": "raised_value_error",
        "exception_type": "ValueError",
        "complete": null,
        "selected_count": 0,
        "review_count": 0,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 0,
        "mock_http_attempts": 0,
        "mock_request_sizes": [],
        "planning_calls": [],
        "plans": [],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": null,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0007248750043800101,
      "returned_context_bytes": 32,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "mock_http_413_loop",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2250,
            "request_json_utf8_bytes": 2250,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002652291994309053,
      "returned_context_bytes": 1108,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "mock_http_413_loop",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2250,
            "request_json_utf8_bytes": 2250,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0022068750113248825,
      "returned_context_bytes": 1108,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "mock_http_413_loop",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2250,
            "request_json_utf8_bytes": 2250,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.005559291996178217,
      "returned_context_bytes": 1108,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "mock_http_413_loop",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2250,
            "request_json_utf8_bytes": 2250,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0024600000033387914,
      "returned_context_bytes": 1108,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "mock_http_413_loop",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2250,
            "request_json_utf8_bytes": 2250,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002182625001296401,
      "returned_context_bytes": 1108,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "mock_http_413_loop",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "loop",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "needs_review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        }
      },
      "result": {
        "status": "needs_review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 4,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 2250,
            "request_json_utf8_bytes": 2250,
            "question_count": 1,
            "criterion_count": 6
          }
        ],
        "planning_calls": [
          {
            "stage": "tool_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 4
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 4
          }
        ],
        "diagnostic": {
          "stage": "tool_selection",
          "reason_codes": [
            "http_413_body_suppressed"
          ],
          "automatic_retry": false
        },
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0022642499970970675,
      "returned_context_bytes": 1108,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    },
    {
      "case": "mock_http_413_direct",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0015105420025065541,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 1
    },
    {
      "case": "mock_http_413_direct",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0019343749881954864,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 1
    },
    {
      "case": "mock_http_413_direct",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.002594749996205792,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 2
    },
    {
      "case": "mock_http_413_direct",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.001359625006443821,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 2
    },
    {
      "case": "mock_http_413_direct",
      "runtime_version": "0.1.9",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0015103339974302799,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "baseline",
      "repeat": 3
    },
    {
      "case": "mock_http_413_direct",
      "runtime_version": "0.1.10",
      "runtime_loaded_from_requested_source": true,
      "fixture": {
        "mode": "direct",
        "mock_response": "http_413"
      },
      "expected": {
        "status": "review",
        "effects": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "usage_complete": false,
        "diagnostic": null
      },
      "result": {
        "status": "review",
        "exception_type": null,
        "complete": false,
        "selected_count": 0,
        "review_count": 2,
        "effects": 0,
        "intent_events": 0,
        "mock_batch_dispatches": 1,
        "mock_http_attempts": 1,
        "mock_request_sizes": [
          {
            "request_json_characters": 638,
            "request_json_utf8_bytes": 638,
            "question_count": 1,
            "criterion_count": 4
          }
        ],
        "planning_calls": [
          {
            "stage": "direct_selection",
            "typed_status_counts": {
              "UNAVAILABLE": 2
            },
            "planning": null,
            "planning_raw_marker_exposures": 0,
            "requests_reported": 1,
            "usage_complete": false,
            "usage_known": false,
            "provider_failure_codes": [
              "http_413_body_suppressed"
            ],
            "raw_planning_matches_expected": true
          }
        ],
        "plans": [
          {
            "items": 1,
            "deferred_status_counts": {},
            "source_count": 2
          }
        ],
        "diagnostic": null,
        "diagnostic_correct": true,
        "planning_correct": true,
        "typed_status_correct": true,
        "behavior_correct": true,
        "usage_complete": false,
        "usage_known": false,
        "reported_mock_usage": null,
        "real_network_attempts": 0,
        "raw_marker_exposures": 0
      },
      "correct": true,
      "whole_operation_seconds": 0.0014559999981429428,
      "returned_context_bytes": 689,
      "actual_model_usage": {
        "requests": 0,
        "models": {},
        "input_tokens": 0,
        "output_tokens": 0,
        "usage_complete": true
      },
      "arm": "candidate",
      "repeat": 3
    }
  ]
}