mcp-security-scanner

Scanning for unprotected MCP servers uses the mcp-security-scanner to detect active endpoints across specified local network ranges. It reports the authentication status of each detected server to flag those lacking proper access controls.

MCP server security auditing detects exposed AI tool endpoints and checks their authentication status. It is needed to identify unprotected servers on your network, preventing unauthorized access to your AI development environment.

Yes, auditing MCP servers requires Node.js version 18 or higher. You also need network access to perform the scan and detect servers across specified network ranges or local hosts.

Yes, configuration auditing checks common AI tool configurations for exposed MCP endpoints. The scanner flags any servers lacking authentication and provides specific remediation advice to secure them.

When the MCP security scanner finds unauthenticated servers, it flags them in a security report and provides remediation advice. This helps you identify exposed endpoints and secure your local AI tools.

The mcp-security-scanner operates using the @contextware/mcp-scan package to perform network scans. It requires Node.js 18+ and network access but lists no additional external dependencies.