Guardrails
Verified IntegrationClient Configuration
— Connect Guardrails to Claude Desktop or Cursor in seconds{
"mcpServers": {
"guardrails-2": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-guardrails-2"
],
"env": {}
}
}
}~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).System Overview
Provides a robust security layer for AI agents and code, implementing comprehensive input validation, output filtering, policy enforcement, audit logging, and rate limiting.
7/23/2026
Open Source
stdio / SSE RPC
Frequently Asked Questions
Architecture and operational details for Guardrails
Yes, Guardrails is designed to act as an MCP (Message Control Protocol) server, making it easily integrable with AI agents like Claude Code, where it intercepts and processes requests and responses to apply its security layers.
Related MCP Servers
Browse all servers →Empower AI assistants to generate and convert documents, manage templates, and automate document workflows efficiently.
Provides a multi-tenant, AI-native Content Delivery Network deployable on Cloudflare, featuring sub-100ms TTFB, AI agent controllability, and comprehensive accessibility.
Deploy a Model Context Protocol server on Cloudflare Workers without requiring authentication.