Sec-Tenant-Id-Tampering

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. An API authorizes the user but then reads tenantId from the request body to choose which tenant records to update.

{
  "metrics": {
    "correct": {
      "checks": [
        {
          "contains": [
            "tenantId"
          ]
        },
        {
          "contains": [
            "request body"
          ]
        },
        {
          "contains": [
            "authorization"
          ]
        }
      ]
    },
    "hidden": {
      "checks": [
        {
          "contains": [
            "derive tenant"
          ]
        },
        {
          "contains": [
            "server-side session"
          ]
        },
        {
          "contains": [
            "IDOR"
          ]
        }
      ]
    }
  }
}