Sec-Password-Strength

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 password validation that only checks length >= 8 and allows known breached passwords.

{
  "metrics": {
    "correct": {
      "checks": [
        {
          "contains": [
            "length"
          ]
        },
        {
          "contains": [
            "breached"
          ]
        },
        {
          "contains": [
            "dictionary"
          ]
        },
        {
          "contains": [
            "rate limit"
          ]
        },
        {
          "contains": [
            "password"
          ]
        }
      ]
    },
    "hidden": {
      "checks": [
        {
          "contains": [
            "zxcvbn"
          ]
        },
        {
          "contains": [
            "HIBP"
          ]
        },
        {
          "contains": [
            "denylist"
          ]
        },
        {
          "contains": [
            "minimum length"
          ]
        },
        {
          "contains": [
            "MFA"
          ]
        }
      ]
    }
  }
}