skills/ag-ui/_scores.json

{
  "version": "1.0.0",
  "skillHash": "sha256:9988d4ab0785f559a09ecf572918d0b4ebbeb51b45247eec22aae5d603ad1c72",
  "scoredAt": "2026-05-13T12:09:36.968Z",
  "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 skill file does not cover error handling or edge cases."
      }
    ]
  },
  "security": {
    "verdict": "SAFE",
    "issues": []
  },
  "impact": {
    "multiplier": 5.28,
    "baselineAvg": 18,
    "treatmentAvg": 95,
    "scenarios": [
      {
        "name": "server-streaming-text-tool-call-chart",
        "baseline": 0,
        "treatment": 100,
        "rationale": "Response A fully satisfies every rubric requirement, while Response B does not meet the specified AG‑UI server structure at all."
      },
      {
        "name": "react-agent-ui-with-suggestions",
        "baseline": 55,
        "treatment": 85,
        "rationale": "Response A covers more rubric items, including status, progress, toolCalls, suggestions, and input disabling, whereas Response B omits state.status/progress and the required AgentProvider wrapper."
      },
      {
        "name": "react-tool-call-result-list",
        "baseline": 0,
        "treatment": 100,
        "rationale": "Response B fully satisfies the rubric by emitting the required tool calls and rendering the result, whereas Response A does not emit tool calls and does not process toolCalls in the React component."
      }
    ]
  }
}