safety_guardrail_breach_analyzer

Active

declared in 0.2.0

Analyzes potential LLM guardrail breaches against IEEE 7000 ethical compliance standards. Designed for risk persona to evaluate safety violations in AI outputs. Accepts raw LLM responses or structured breach reports, returns compliance analysis with severity scoring and mitigation recommendations.

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"
    }
  }
}

What this tool wraps· 0 endpoints

min confidence0.70 0.50

No endpoints wrapped at confidence ≥ 0.50.

Parent server

@gapup/mcp-knowledge

https://github.com/getgapup/gapup-mcp

2/7 registries

View full server →