Refactor-Permission-Guards

The model receives the prompt (and optional system message). The run uses scorer rubric_json_metrics with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.

Return JSON only with keys plan, refactor, tests. A controller has nested if statements checking role, ownership, and feature flags before every action. Refactor the guards for clarity.

{
  "metrics": {
    "intent": {
      "checks": [
        {
          "contains": [
            "nested if"
          ]
        },
        {
          "contains": [
            "guard clauses"
          ]
        },
        {
          "contains": [
            "permissions"
          ]
        }
      ]
    },
    "visible": {
      "checks": [
        {
          "contains": [
            "role"
          ]
        },
        {
          "contains": [
            "ownership"
          ]
        },
        {
          "contains": [
            "feature flags"
          ]
        }
      ]
    },
    "hidden": {
      "checks": [
        {
          "contains": [
            "same authorization"
          ]
        },
        {
          "contains": [
            "negative tests"
          ]
        },
        {
          "contains": [
            "clarity"
          ]
        }
      ]
    }
  }
}