skills/activepieces/_scores.json

{
  "version": "1.0.0",
  "skillHash": "sha256:798a3823b9e1e04a89edbbbfd88ba5c18a435571e69cd0af9d12fc0b4253324c",
  "scoredAt": "2026-05-13T11:44:43.769Z",
  "backend": "ollama",
  "model": "gpt-oss:20b",
  "quality": {
    "score": 93,
    "dimensions": {
      "clarity": "PASS",
      "completeness": "WEAK",
      "conciseness": "PASS",
      "actionability": "PASS",
      "crossPlatform": "PASS",
      "examples": "PASS"
    },
    "issues": [
      {
        "severity": "MEDIUM",
        "category": "completeness",
        "detail": "The instructions do not cover error handling or edge cases."
      }
    ]
  },
  "security": {
    "verdict": "SAFE",
    "issues": []
  },
  "impact": {
    "multiplier": 4.1,
    "baselineAvg": 20,
    "treatmentAvg": 82,
    "scenarios": [
      {
        "name": "self-host-activepieces-docker-compose",
        "baseline": 0,
        "treatment": 100,
        "rationale": "Response A fully satisfies the rubric, while Response B omits required variables, port mapping, and volume naming."
      },
      {
        "name": "create-order-processing-code-step",
        "baseline": 40,
        "treatment": 100,
        "rationale": "Response A fully satisfies the rubric, while Response B deviates in naming, type, and output structure."
      },
      {
        "name": "build-custom-piece-my-app",
        "baseline": 20,
        "treatment": 45,
        "rationale": "Response B uses createPiece (though not PieceAuth.SecretText) whereas Response A uses registerPiece, so B covers more rubric items."
      }
    ]
  }
}