{
  "cells": [
    {
      "skill": "code-review-and-quality",
      "A": "claude-opus-5",
      "B": "claude-opus-5-5",
      "columns": {
        "original": {
          "labels": [
            "no separation detected"
          ],
          "cases": [
            {
              "id": "severity-labeled-findings",
              "A": {
                "mean": 0.901,
                "sd": 0.007
              },
              "B": {
                "mean": 0.91,
                "sd": 0.015
              },
              "delta": 0.009,
              "label": "no separation detected"
            }
          ]
        },
        "control": {
          "labels": [
            "no separation detected"
          ],
          "cases": [
            {
              "id": "severity-labeled-findings",
              "A": {
                "mean": 0.899,
                "sd": 0.005
              },
              "B": {
                "mean": 0.902,
                "sd": 0.016
              },
              "delta": 0.003,
              "label": "no separation detected"
            }
          ]
        },
        "treatment": {
          "labels": [
            "no separation detected"
          ],
          "cases": [
            {
              "id": "severity-labeled-findings",
              "A": {
                "mean": 0.901,
                "sd": 0.013
              },
              "B": {
                "mean": 0.912,
                "sd": 0.012
              },
              "delta": 0.011,
              "label": "no separation detected"
            }
          ]
        }
      }
    },
    {
      "skill": "git-workflow-and-versioning",
      "A": "claude-opus-5",
      "B": "claude-opus-5-5",
      "columns": {
        "original": {
          "labels": [
            "no separation detected; Not enough draws to conclude at this effect floor"
          ],
          "cases": [
            {
              "id": "commit-message-conventional-type",
              "A": {
                "mean": 0.861,
                "sd": 0.008
              },
              "B": {
                "mean": 0.838,
                "sd": 0.027
              },
              "delta": -0.023,
              "label": "no separation detected; Not enough draws to conclude at this effect floor"
            }
          ]
        },
        "control": {
          "labels": [
            "no separation detected"
          ],
          "cases": [
            {
              "id": "commit-message-conventional-type",
              "A": {
                "mean": 0.853,
                "sd": 0.007
              },
              "B": {
                "mean": 0.844,
                "sd": 0.015
              },
              "delta": -0.009,
              "label": "no separation detected"
            }
          ]
        },
        "treatment": {
          "labels": [
            "no separation detected"
          ],
          "cases": [
            {
              "id": "commit-message-conventional-type",
              "A": {
                "mean": 0.866,
                "sd": 0.005
              },
              "B": {
                "mean": 0.858,
                "sd": 0.01
              },
              "delta": -0.008,
              "label": "no separation detected"
            }
          ]
        }
      }
    },
    {
      "skill": "documentation-and-adrs",
      "A": "claude-opus-5",
      "B": "claude-opus-5-5",
      "columns": {
        "original": {
          "labels": [
            "no separation detected; Not enough draws to conclude at this effect floor"
          ],
          "cases": [
            {
              "id": "comment-intent-not-implementation",
              "A": {
                "mean": 0.804,
                "sd": 0.103
              },
              "B": {
                "mean": 0.661,
                "sd": 0.148
              },
              "delta": -0.143,
              "label": "no separation detected; Not enough draws to conclude at this effect floor"
            }
          ]
        },
        "control": {
          "labels": [
            "no separation detected; Not enough draws to conclude at this effect floor"
          ],
          "cases": [
            {
              "id": "comment-intent-not-implementation",
              "A": {
                "mean": 0.807,
                "sd": 0.109
              },
              "B": {
                "mean": 0.69,
                "sd": 0.147
              },
              "delta": -0.117,
              "label": "no separation detected; Not enough draws to conclude at this effect floor"
            }
          ]
        },
        "treatment": {
          "labels": [
            "no separation detected; Not enough draws to conclude at this effect floor"
          ],
          "cases": [
            {
              "id": "comment-intent-not-implementation",
              "A": {
                "mean": 0.801,
                "sd": 0.099
              },
              "B": {
                "mean": 0.658,
                "sd": 0.105
              },
              "delta": -0.144,
              "label": "no separation detected; Not enough draws to conclude at this effect floor"
            }
          ]
        }
      }
    }
  ],
  "receipts": {
    "original": {
      "code-review-and-quality": {
        "claude-opus-5": {
          "with_skill": 0.901,
          "baseline": 0.862,
          "lift": 0.039,
          "judge": "claude-opus-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.91,
          "baseline": 0.68,
          "lift": 0.23,
          "judge": "claude-opus-5",
          "level": "TESTED"
        }
      },
      "git-workflow-and-versioning": {
        "claude-opus-5": {
          "with_skill": 0.861,
          "baseline": 0.765,
          "lift": 0.096,
          "judge": "claude-opus-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.838,
          "baseline": 0.3,
          "lift": 0.538,
          "judge": "claude-opus-5",
          "level": "TESTED"
        }
      },
      "documentation-and-adrs": {
        "claude-opus-5": {
          "with_skill": 0.804,
          "baseline": 0.567,
          "lift": 0.237,
          "judge": "claude-opus-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.661,
          "baseline": 0.567,
          "lift": 0.094,
          "judge": "claude-opus-5",
          "level": "TESTED"
        }
      }
    },
    "control": {
      "code-review-and-quality": {
        "claude-opus-5": {
          "with_skill": 0.899,
          "baseline": 0.87,
          "lift": 0.029,
          "judge": "claude-opus-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.902,
          "baseline": 0.693,
          "lift": 0.209,
          "judge": "claude-opus-5",
          "level": "TESTED"
        }
      },
      "git-workflow-and-versioning": {
        "claude-opus-5": {
          "with_skill": 0.853,
          "baseline": 0.766,
          "lift": 0.087,
          "judge": "claude-opus-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.844,
          "baseline": 0.3,
          "lift": 0.544,
          "judge": "claude-opus-5",
          "level": "TESTED"
        }
      },
      "documentation-and-adrs": {
        "claude-opus-5": {
          "with_skill": 0.807,
          "baseline": 0.583,
          "lift": 0.223,
          "judge": "claude-opus-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.69,
          "baseline": 0.573,
          "lift": 0.117,
          "judge": "claude-opus-5",
          "level": "TESTED"
        }
      }
    },
    "treatment": {
      "code-review-and-quality": {
        "claude-opus-5": {
          "with_skill": 0.901,
          "baseline": 0.788,
          "lift": 0.113,
          "judge": "claude-opus-5-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.912,
          "baseline": 0.603,
          "lift": 0.309,
          "judge": "claude-opus-5-5",
          "level": "TESTED"
        }
      },
      "git-workflow-and-versioning": {
        "claude-opus-5": {
          "with_skill": 0.866,
          "baseline": 0.774,
          "lift": 0.091,
          "judge": "claude-opus-5-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.858,
          "baseline": 0.3,
          "lift": 0.558,
          "judge": "claude-opus-5-5",
          "level": "TESTED"
        }
      },
      "documentation-and-adrs": {
        "claude-opus-5": {
          "with_skill": 0.801,
          "baseline": 0.517,
          "lift": 0.285,
          "judge": "claude-opus-5-5",
          "level": "TESTED"
        },
        "claude-opus-5-5": {
          "with_skill": 0.658,
          "baseline": 0.587,
          "lift": 0.071,
          "judge": "claude-opus-5-5",
          "level": "TESTED"
        }
      }
    }
  },
  "agreement": {
    "treatment_vs_control": {
      "n": 51,
      "same_side": 50,
      "mean_abs_diff": 0.0291
    },
    "control_vs_original": {
      "n": 51,
      "same_side": 50,
      "mean_abs_diff": 0.0127
    }
  }
}
