safety_guardrail_breach_analyzer

Parameters schema

{
  "type": "object",
  "required": [
    "llmOutput"
  ],
  "properties": {
    "async": {
      "type": "boolean",
      "description": "If true, returns a job_id immediately (<200ms) instead of waiting for the result. Poll the result with job_result(job_id). Use for slow tools to avoid client timeouts."
    },
    "context": {
      "type": "string",
      "description": "Contextual information about the prompt or conversation"
    },
    "llmOutput": {
      "type": "string",
      "description": "Raw text output from LLM to analyze for guardrail breaches"
    },
    "severityThreshold": {
      "type": "number",
      "default": 5,
      "maximum": 10,
      "minimum": 0,
      "description": "Minimum severity score to report (0-10 scale)"
    },
    "includeMitigations": {
      "type": "boolean",
      "default": true,
      "description": "Whether to include mitigation recommendations"
    }
  }
}

No endpoints wrapped at confidence ≥ 0.70.

Parent server