Agent Readiness Auditor

Overview

Agent Readiness Auditor is a Model Context Protocol server that evaluates websites for AI agent readiness and safety. It scores prompt injection vulnerabilities and checks llms.txt compliance, giving AI assistants direct access to comprehensive security auditing capabilities that integrate seamlessly into agent workflows.

The server runs locally via npx -y agent-readiness-auditor with no authentication required. Built with TypeScript and published as v0.4.7 under the MIT license, Agent Readiness Auditor is maintained at asish-singh/agent-readiness-auditor on GitHub and operates over stdio on your machine.

Agent Readiness Auditor is categorized under AI & Agents and works with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and Zed. Maintainers can claim this listing at mcp.site to verify ownership and keep information current.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "agent_readiness_auditor": {
      "command": "npx",
      "args": [
        "-y",
        "agent-readiness-auditor"
      ]
    }
  }
}