{
  "timestamp": "2026-11-10T12:40:28.042Z",
  "model": "claude-direct/claude-sonnet-5-5",
  "scenarios": [
    {
      "scenario": "scenario-A-noisy-folder",
      "model": "claude-direct/claude-sonnet-5-5",
      "mode": "cloud",
      "timestamp": "2026-11-10T12:40:28.038Z",
      "results": {
        "claude-direct": {
          "no-memory": {
            "tokens": {
              "input": 129,
              "output": 1787,
              "input_cached": 0
            },
            "time_ms": 12642,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 1916,
            "token_samples": [
              2439,
              1546,
              2100,
              2177,
              1575,
              1742,
              1631,
              2162,
              1904,
              1916
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0
              },
              {
                "n": 2,
                "pass_rate": 0
              },
              {
                "n": 3,
                "pass_rate": 0
              },
              {
                "n": 4,
                "pass_rate": 0
              },
              {
                "n": 5,
                "pass_rate": 0
              },
              {
                "n": 6,
                "pass_rate": 0
              },
              {
                "n": 7,
                "pass_rate": 0
              }
            ]
          },
          "oracle-ceiling": {
            "tokens": {
              "input": 674,
              "output": 3773,
              "input_cached": 0
            },
            "time_ms": 23197,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 4447,
            "token_samples": [
              4478,
              4314,
              4335,
              4201,
              4473,
              4256,
              4239,
              4522,
              4487,
              4447
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 4375,
            "retrieval": {
              "recall": {
                "1": 0.333,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0.9
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "keyword": {
            "tokens": {
              "input": 476,
              "output": 2156,
              "input_cached": 0
            },
            "time_ms": 14832,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 2632,
            "token_samples": [
              2632,
              2691,
              2541,
              2539,
              2428,
              2485,
              2704,
              2710,
              2848,
              2623
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": {
              "recall": {
                "1": 0.333,
                "3": 0.333,
                "5": 0.333,
                "10": 0.333
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0
              },
              {
                "n": 2,
                "pass_rate": 0
              },
              {
                "n": 3,
                "pass_rate": 0
              },
              {
                "n": 4,
                "pass_rate": 0.1
              },
              {
                "n": 5,
                "pass_rate": 0
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "vector-baseline": {
            "tokens": {
              "input": 394,
              "output": 2110,
              "input_cached": 0
            },
            "time_ms": 13086,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 2504,
            "token_samples": [
              2584,
              2593,
              2603,
              3058,
              2300,
              2345,
              2504,
              2274,
              2428,
              2389
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": {
              "recall": {
                "1": 0,
                "3": 0,
                "5": 0,
                "10": 0
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0
              },
              {
                "n": 2,
                "pass_rate": 0
              },
              {
                "n": 3,
                "pass_rate": 0
              },
              {
                "n": 4,
                "pass_rate": 0
              },
              {
                "n": 5,
                "pass_rate": 0
              },
              {
                "n": 6,
                "pass_rate": 0
              },
              {
                "n": 7,
                "pass_rate": 0
              }
            ]
          },
          "context-dump-bounded": {
            "tokens": {
              "input": 11098,
              "output": 4216,
              "input_cached": 0
            },
            "time_ms": 28430,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 15314,
            "token_samples": [
              16424,
              15185,
              15600,
              15279,
              15921,
              15252,
              15024,
              15220,
              15314,
              15434
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 15465,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0.8
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "context-dump-large": {
            "tokens": {
              "input": 80730,
              "output": 4405,
              "input_cached": 0
            },
            "time_ms": 27996,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 85135,
            "token_samples": [
              84593,
              85278,
              85135,
              84863,
              85829,
              85588,
              85977,
              85061,
              85100,
              85039
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 85246,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0.9
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "brain-full": {
            "tokens": {
              "input": 817,
              "output": 2960,
              "input_cached": 0
            },
            "time_ms": 19475,
            "runs": 10,
            "total": 10,
            "passes": 9,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0.9,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 0.9,
            "judge_pass_rate": 0.9,
            "median_tokens": 3777,
            "token_samples": [
              3777,
              4106,
              3793,
              3658,
              4394,
              3750,
              3739,
              3616,
              3689,
              4003
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 4281,
            "retrieval": {
              "recall": {
                "1": 0.333,
                "3": 0.667,
                "5": 0.667,
                "10": 0.667
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0.9
              },
              {
                "n": 4,
                "pass_rate": 0
              },
              {
                "n": 5,
                "pass_rate": 0
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "brain-connector-gated": {
            "tokens": {
              "input": 1260,
              "output": 3687,
              "input_cached": 0
            },
            "time_ms": 23790,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 4947,
            "token_samples": [
              4635,
              5064,
              4947,
              4835,
              5072,
              4817,
              4694,
              4868,
              5129,
              5097
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 4916,
            "retrieval": {
              "recall": {
                "1": 0.333,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "brain-no-recall": {
            "tokens": {
              "input": 377,
              "output": 3458,
              "input_cached": 0
            },
            "time_ms": 22627,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 3835,
            "token_samples": [
              3510,
              3897,
              3786,
              3835,
              3605,
              4093,
              3545,
              3909,
              3599,
              4259
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 3804,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0.6
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 0
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          },
          "brain-no-pin": {
            "tokens": {
              "input": 817,
              "output": 3053,
              "input_cached": 0
            },
            "time_ms": 20119,
            "runs": 10,
            "total": 10,
            "passes": 8,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0.8,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 0.8,
            "judge_pass_rate": 0.8,
            "median_tokens": 3870,
            "token_samples": [
              3825,
              4045,
              3647,
              3842,
              3885,
              3695,
              3870,
              4020,
              3684,
              4102
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 4827,
            "retrieval": {
              "recall": {
                "1": 0.333,
                "3": 0.667,
                "5": 0.667,
                "10": 0.667
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0.8
              },
              {
                "n": 4,
                "pass_rate": 0
              },
              {
                "n": 5,
                "pass_rate": 0
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              }
            ]
          }
        }
      },
      "merge": {
        "sources": {
          "no-memory": "results/_sonnet-campaign-20261110/partial",
          "oracle-ceiling": "results/_sonnet-campaign-20261110/partial",
          "keyword": "results/_sonnet-campaign-20261110/partial",
          "vector-baseline": "results/_sonnet-campaign-20261110/partial",
          "context-dump-bounded": "results/_sonnet-campaign-20261110/partial-resume",
          "context-dump-large": "results/_sonnet-campaign-20261110/partial-resume",
          "brain-full": "results/_sonnet-campaign-20261110/partial-resume",
          "brain-connector-gated": "results/_sonnet-campaign-20261110/partial-fix",
          "brain-no-recall": "results/_sonnet-campaign-20261110/partial-resume",
          "brain-no-pin": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "superseded": {
          "brain-connector-gated": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "missing_arms": [],
        "excluded_arms": [
          "dense"
        ]
      }
    },
    {
      "scenario": "scenario-B-three-sessions",
      "model": "claude-direct/claude-sonnet-5-5",
      "mode": "cloud",
      "timestamp": "2026-11-10T12:40:28.040Z",
      "results": {
        "claude-direct": {
          "fixture-only": {
            "tokens": {
              "input": 89,
              "output": 905,
              "input_cached": 0
            },
            "time_ms": 8448,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 994,
            "token_samples": [
              974,
              994,
              844,
              931,
              902,
              1855,
              2032,
              1060,
              1833,
              984
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0.3
              },
              {
                "n": 2,
                "pass_rate": 0
              },
              {
                "n": 3,
                "pass_rate": 0.3
              },
              {
                "n": 4,
                "pass_rate": 0.3
              },
              {
                "n": 5,
                "pass_rate": 0.7
              },
              {
                "n": 6,
                "pass_rate": 0
              }
            ]
          },
          "oracle-ceiling": {
            "tokens": {
              "input": 357,
              "output": 2019,
              "input_cached": 0
            },
            "time_ms": 13169,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2376,
            "token_samples": [
              2505,
              2206,
              2376,
              2295,
              2438,
              2144,
              2389,
              2361,
              2557,
              2366
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2364,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "keyword": {
            "tokens": {
              "input": 436,
              "output": 1953,
              "input_cached": 0
            },
            "time_ms": 13824,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2389,
            "token_samples": [
              2152,
              2392,
              2528,
              2319,
              2389,
              2270,
              2481,
              2457,
              2388,
              2263
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2364,
            "retrieval": {
              "recall": {
                "1": 0,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-real": {
            "tokens": {
              "input": 851,
              "output": 2322,
              "input_cached": 0
            },
            "time_ms": 16841,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 3173,
            "token_samples": [
              3468,
              3629,
              3372,
              3124,
              3367,
              2868,
              3080,
              3173,
              3107,
              2885
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 3207,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-connector-gated": {
            "tokens": {
              "input": 1009,
              "output": 2200,
              "input_cached": 0
            },
            "time_ms": 15297,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 3209,
            "token_samples": [
              3317,
              3520,
              3389,
              3630,
              2913,
              2907,
              3209,
              2899,
              3127,
              2868
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 3178,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-no-pin": {
            "tokens": {
              "input": 851,
              "output": 2295,
              "input_cached": 0
            },
            "time_ms": 15620,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 3146,
            "token_samples": [
              3005,
              2999,
              3117,
              3498,
              3004,
              2889,
              3534,
              3263,
              3505,
              3146
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 3196,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-no-recall": {
            "tokens": {
              "input": 338,
              "output": 1862,
              "input_cached": 0
            },
            "time_ms": 13159,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2200,
            "token_samples": [
              2301,
              2156,
              2215,
              2028,
              2200,
              2129,
              2112,
              2142,
              2383,
              2428
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2209,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          }
        }
      },
      "merge": {
        "sources": {
          "fixture-only": "results/_sonnet-campaign-20261110/partial",
          "oracle-ceiling": "results/_sonnet-campaign-20261110/partial",
          "keyword": "results/_sonnet-campaign-20261110/partial-resume",
          "brain-real": "results/_sonnet-campaign-20261110/partial-resume",
          "brain-connector-gated": "results/_sonnet-campaign-20261110/partial-fix",
          "brain-no-pin": "results/_sonnet-campaign-20261110/partial-resume",
          "brain-no-recall": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "superseded": {
          "brain-connector-gated": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "missing_arms": [],
        "excluded_arms": [
          "dense"
        ]
      }
    },
    {
      "scenario": "scenario-C-contradiction",
      "model": "claude-direct/claude-sonnet-5-5",
      "mode": "cloud",
      "timestamp": "2026-11-10T12:40:28.041Z",
      "results": {
        "claude-direct": {
          "fixture-only": {
            "tokens": {
              "input": 98,
              "output": 705,
              "input_cached": 0
            },
            "time_ms": 5784,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 803,
            "token_samples": [
              877,
              794,
              751,
              761,
              763,
              813,
              805,
              856,
              769,
              803
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0
              },
              {
                "n": 2,
                "pass_rate": 0
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 0
              }
            ]
          },
          "oracle-ceiling": {
            "tokens": {
              "input": 285,
              "output": 555,
              "input_cached": 0
            },
            "time_ms": 5090,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 840,
            "token_samples": [
              870,
              776,
              739,
              810,
              840,
              863,
              857,
              766,
              832,
              857
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 821,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "keyword": {
            "tokens": {
              "input": 396,
              "output": 515,
              "input_cached": 0
            },
            "time_ms": 4192,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 911,
            "token_samples": [
              847,
              941,
              1021,
              894,
              843,
              916,
              911,
              942,
              889,
              908
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 911,
            "retrieval": {
              "recall": {
                "1": 0,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-real": {
            "tokens": {
              "input": 643,
              "output": 513,
              "input_cached": 0
            },
            "time_ms": 4217,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 1156,
            "token_samples": [
              1129,
              1131,
              1156,
              1128,
              1120,
              1169,
              1122,
              1170,
              1173,
              1185
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1148,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-no-pin": {
            "tokens": {
              "input": 643,
              "output": 527,
              "input_cached": 0
            },
            "time_ms": 4468,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 1170,
            "token_samples": [
              1210,
              1145,
              1130,
              1217,
              1109,
              1209,
              1170,
              1151,
              1121,
              1181
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1164,
            "retrieval": {
              "recall": {
                "1": 1,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "dump-all-chrono": {
            "tokens": {
              "input": 397,
              "output": 412,
              "input_cached": 0
            },
            "time_ms": 3645,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 809,
            "token_samples": [
              719,
              861,
              821,
              797,
              898,
              763,
              888,
              789,
              804,
              809
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 815,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          }
        }
      },
      "merge": {
        "sources": {
          "fixture-only": "results/_sonnet-campaign-20261110/partial",
          "oracle-ceiling": "results/_sonnet-campaign-20261110/partial",
          "keyword": "results/_sonnet-campaign-20261110/partial",
          "brain-real": "results/_sonnet-campaign-20261110/partial",
          "brain-no-pin": "results/_sonnet-campaign-20261110/partial",
          "dump-all-chrono": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "superseded": {},
        "missing_arms": [],
        "excluded_arms": [
          "dense"
        ]
      }
    },
    {
      "scenario": "scenario-D-skills",
      "model": "claude-direct/claude-sonnet-5-5",
      "mode": "cloud",
      "timestamp": "2026-11-10T12:40:28.041Z",
      "results": {
        "claude-direct": {
          "fixture-only": {
            "tokens": {
              "input": 85,
              "output": 2522,
              "input_cached": 0
            },
            "time_ms": 16865,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 2607,
            "token_samples": [
              2156,
              2471,
              3101,
              2548,
              2607,
              2332,
              2879,
              2947,
              3007,
              1584
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0
              },
              {
                "n": 2,
                "pass_rate": 0
              },
              {
                "n": 3,
                "pass_rate": 0.1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 0.9
              },
              {
                "n": 6,
                "pass_rate": 0
              },
              {
                "n": 7,
                "pass_rate": 0.9
              },
              {
                "n": 8,
                "pass_rate": 0.9
              }
            ]
          },
          "brain-skills-L0": {
            "tokens": {
              "input": 85,
              "output": 2448,
              "input_cached": 0
            },
            "time_ms": 16658,
            "runs": 10,
            "total": 10,
            "passes": 1,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0.1,
            "no_completion_rate": 0,
            "success": false,
            "consistency": 0.1,
            "judge_pass_rate": 0.1,
            "median_tokens": 2533,
            "token_samples": [
              2335,
              2416,
              2717,
              2533,
              3448,
              1399,
              1444,
              2893,
              2606,
              2337
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 24128,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0
              },
              {
                "n": 2,
                "pass_rate": 0.3
              },
              {
                "n": 3,
                "pass_rate": 0.1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 0
              },
              {
                "n": 7,
                "pass_rate": 0.7
              },
              {
                "n": 8,
                "pass_rate": 0.8
              }
            ]
          },
          "brain-skills-loaded": {
            "tokens": {
              "input": 632,
              "output": 1740,
              "input_cached": 0
            },
            "time_ms": 11315,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2372,
            "token_samples": [
              2343,
              2372,
              2332,
              2295,
              2421,
              2399,
              2376,
              2359,
              2279,
              2398
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2357,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0.8
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              },
              {
                "n": 8,
                "pass_rate": 1
              }
            ]
          },
          "brain-skills-all-loaded": {
            "tokens": {
              "input": 888,
              "output": 2416,
              "input_cached": 0
            },
            "time_ms": 17304,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 3304,
            "token_samples": [
              3084,
              3102,
              3666,
              3296,
              3344,
              3051,
              3618,
              3146,
              3304,
              3321
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 3293,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              },
              {
                "n": 7,
                "pass_rate": 1
              },
              {
                "n": 8,
                "pass_rate": 1
              }
            ]
          }
        }
      },
      "merge": {
        "sources": {
          "fixture-only": "results/_sonnet-campaign-20261110/partial",
          "brain-skills-L0": "results/_sonnet-campaign-20261110/partial",
          "brain-skills-loaded": "results/_sonnet-campaign-20261110/partial",
          "brain-skills-all-loaded": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "superseded": {},
        "missing_arms": [],
        "excluded_arms": [
          "dense"
        ]
      }
    },
    {
      "scenario": "scenario-F-abstention",
      "model": "claude-direct/claude-sonnet-5-5",
      "mode": "cloud",
      "timestamp": "2026-11-10T12:40:28.041Z",
      "results": {
        "claude-direct": {
          "fixture-only": {
            "tokens": {
              "input": 140,
              "output": 866,
              "input_cached": 0
            },
            "time_ms": 8244,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 1006,
            "token_samples": [
              1040,
              964,
              1125,
              1056,
              977,
              932,
              910,
              1062,
              985,
              1006
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1006,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "oracle-ceiling": {
            "tokens": {
              "input": 387,
              "output": 988,
              "input_cached": 0
            },
            "time_ms": 8617,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 1375,
            "token_samples": [
              1324,
              1260,
              1229,
              1400,
              1435,
              1422,
              1332,
              1412,
              1195,
              1375
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1338,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "keyword": {
            "tokens": {
              "input": 392,
              "output": 929,
              "input_cached": 0
            },
            "time_ms": 8013,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 1321,
            "token_samples": [
              1326,
              1295,
              1310,
              1321,
              1274,
              1279,
              1261,
              1402,
              1458,
              1349
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1328,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          },
          "brain-real": {
            "tokens": {
              "input": 720,
              "output": 1061,
              "input_cached": 0
            },
            "time_ms": 9638,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 1781,
            "token_samples": [
              1781,
              1731,
              1840,
              1831,
              1773,
              1772,
              1781,
              1767,
              1751,
              1809
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1784,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              },
              {
                "n": 6,
                "pass_rate": 1
              }
            ]
          }
        }
      },
      "merge": {
        "sources": {
          "fixture-only": "results/_sonnet-campaign-20261110/partial",
          "oracle-ceiling": "results/_sonnet-campaign-20261110/partial",
          "keyword": "results/_sonnet-campaign-20261110/partial",
          "brain-real": "results/_sonnet-campaign-20261110/partial"
        },
        "superseded": {},
        "missing_arms": [],
        "excluded_arms": [
          "dense"
        ]
      }
    },
    {
      "scenario": "scenario-G-poisoned-haystack",
      "model": "claude-direct/claude-sonnet-5-5",
      "mode": "cloud",
      "timestamp": "2026-11-10T12:40:28.042Z",
      "results": {
        "claude-direct": {
          "fixture-only": {
            "tokens": {
              "input": 112,
              "output": 1377,
              "input_cached": 0
            },
            "time_ms": 11150,
            "runs": 10,
            "total": 10,
            "passes": 9,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 0.9,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 0.9,
            "judge_pass_rate": 0.9,
            "median_tokens": 1489,
            "token_samples": [
              1754,
              1231,
              1671,
              1520,
              1489,
              1447,
              1870,
              1346,
              1341,
              1094
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 1640,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 0.9
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 0
              }
            ]
          },
          "oracle-ceiling": {
            "tokens": {
              "input": 444,
              "output": 1635,
              "input_cached": 0
            },
            "time_ms": 13385,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2079,
            "token_samples": [
              2239,
              1999,
              1983,
              2120,
              2079,
              2187,
              2019,
              1962,
              2031,
              2384
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2100,
            "retrieval": {
              "recall": {
                "1": 0.5,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 0
              }
            ]
          },
          "brain-flag": {
            "tokens": {
              "input": 769,
              "output": 1664,
              "input_cached": 0
            },
            "time_ms": 12916,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2433,
            "token_samples": [
              2394,
              2433,
              2433,
              2373,
              2383,
              2501,
              2444,
              2459,
              2425,
              2400
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2425,
            "retrieval": {
              "recall": {
                "1": 0.5,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 0
              }
            ]
          },
          "brain-enforce": {
            "tokens": {
              "input": 769,
              "output": 1548,
              "input_cached": 0
            },
            "time_ms": 11614,
            "runs": 10,
            "total": 10,
            "passes": 10,
            "completed": 10,
            "no_completion": 0,
            "completion_rate": 1,
            "success_rate": 1,
            "no_completion_rate": 0,
            "success": true,
            "consistency": 1,
            "judge_pass_rate": 1,
            "median_tokens": 2317,
            "token_samples": [
              2283,
              2389,
              2396,
              2230,
              2317,
              2267,
              2420,
              2282,
              2349,
              2262
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 2320,
            "retrieval": {
              "recall": {
                "1": 0.5,
                "3": 1,
                "5": 1,
                "10": 1
              }
            },
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 1
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 0
              }
            ]
          },
          "brain-no-quarantine": {
            "tokens": {
              "input": 0,
              "output": 0,
              "input_cached": 0
            },
            "time_ms": 1458,
            "runs": 10,
            "total": 10,
            "passes": 0,
            "completed": 0,
            "no_completion": 10,
            "completion_rate": 0,
            "success_rate": 0,
            "no_completion_rate": 1,
            "success": false,
            "consistency": 0,
            "judge_pass_rate": 0,
            "median_tokens": 0,
            "token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": null,
            "retrieval": null,
            "criteria_pass_rate": null
          },
          "context-dump-bounded": {
            "tokens": {
              "input": 9225,
              "output": 1620,
              "input_cached": 0
            },
            "time_ms": 12948,
            "runs": 10,
            "total": 10,
            "passes": 9,
            "completed": 9,
            "no_completion": 1,
            "completion_rate": 0.9,
            "success_rate": 0.9,
            "no_completion_rate": 0.1,
            "success": true,
            "consistency": 0.9,
            "judge_pass_rate": 0.9,
            "median_tokens": 10845,
            "token_samples": [
              10689,
              10858,
              10736,
              10857,
              10997,
              10721,
              10748,
              10845,
              10889
            ],
            "cached_token_samples": [
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0,
              0
            ],
            "tokens_per_success": 10816,
            "retrieval": null,
            "criteria_pass_rate": [
              {
                "n": 1,
                "pass_rate": 1
              },
              {
                "n": 2,
                "pass_rate": 1
              },
              {
                "n": 3,
                "pass_rate": 0
              },
              {
                "n": 4,
                "pass_rate": 1
              },
              {
                "n": 5,
                "pass_rate": 1
              }
            ]
          }
        }
      },
      "merge": {
        "sources": {
          "fixture-only": "results/_sonnet-campaign-20261110/partial",
          "oracle-ceiling": "results/_sonnet-campaign-20261110/partial",
          "brain-flag": "results/_sonnet-campaign-20261110/partial",
          "brain-enforce": "results/_sonnet-campaign-20261110/partial",
          "brain-no-quarantine": "results/_sonnet-campaign-20261110/partial-fix",
          "context-dump-bounded": "results/_sonnet-campaign-20261110/partial-fix"
        },
        "superseded": {
          "brain-no-quarantine": "results/_sonnet-campaign-20261110/partial-resume",
          "context-dump-bounded": "results/_sonnet-campaign-20261110/partial-resume"
        },
        "missing_arms": [],
        "excluded_arms": [
          "dense"
        ]
      }
    }
  ]
}