"author:fpytloun" matching MCP servers:

Guardrails service for AI agents that evaluates every tool call for safety and alignment before execution, providing default-deny policy, LLM safety evaluation, and audit trail.