skills/ai-content-detection/_scores.json

{
  "version": "1.0.0",
  "skillHash": "sha256:f1b7e0fd5e0a2251a67b3d9c8addc0379fb4a48e2c83553bc7fb9909f667bd92",
  "scoredAt": "2026-05-13T13:39:48.964Z",
  "backend": "ollama",
  "model": "gpt-oss:20b",
  "quality": {
    "score": 80,
    "dimensions": {
      "clarity": "PASS",
      "completeness": "WEAK",
      "conciseness": "PASS",
      "actionability": "WEAK",
      "crossPlatform": "WEAK",
      "examples": "PASS"
    },
    "issues": [
      {
        "severity": "MEDIUM",
        "category": "completeness",
        "detail": "The skill does not specify how to handle API failures or malformed responses, leaving error handling undefined."
      },
      {
        "severity": "MEDIUM",
        "category": "actionability",
        "detail": "The instructions lack concrete code or example prompts, making it difficult to implement the detection pipeline directly."
      },
      {
        "severity": "MEDIUM",
        "category": "crossPlatform",
        "detail": "The skill requires Node.js 18+ and the Anthropic SDK, limiting compatibility to specific environments."
      }
    ]
  },
  "security": {
    "verdict": "SAFE",
    "issues": []
  },
  "impact": {
    "multiplier": 0,
    "baselineAvg": 0,
    "treatmentAvg": 0,
    "scenarios": [
      {
        "name": "short-text-skip",
        "baseline": 0,
        "treatment": 0,
        "rationale": "Both responses omit the required statement about skipping automated scoring for texts under 200 words and the explanation of why short texts yield unreliable results."
      },
      {
        "name": "full-detection-pipeline",
        "baseline": 0,
        "treatment": 0,
        "rationale": "Neither response provides the required analysis results, verdicts, or scores; both fall short of the rubric."
      }
    ]
  }
}