Sec-Insecure-Config-Scanner

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 verdict, risk, mitigation. Review a config scanner that must detect debug mode, permissive CORS, weak TLS, and public admin endpoints.

{
  "metrics": {
    "correct": {
      "checks": [
        {
          "contains": [
            "debug"
          ]
        },
        {
          "contains": [
            "CORS"
          ]
        },
        {
          "contains": [
            "TLS"
          ]
        },
        {
          "contains": [
            "admin"
          ]
        },
        {
          "contains": [
            "config"
          ]
        }
      ]
    },
    "hidden": {
      "checks": [
        {
          "contains": [
            "wildcard"
          ]
        },
        {
          "contains": [
            "production"
          ]
        },
        {
          "contains": [
            "HSTS"
          ]
        },
        {
          "contains": [
            "secret"
          ]
        },
        {
          "contains": [
            "deny"
          ]
        }
      ]
    }
  }
}