ABOUT

About MCP Guard

MCP Guard is an independent, free web project built to find risky commands, exposed secrets, and unsafe patterns in MCP configuration. It is designed for developers and teams reviewing local MCP server configurations before use. The tool runs without an account and aims to turn a narrow technical task into a clear, repeatable workflow.

What this site provides

Checks are paired with plain-English remediation guidance. The scanner is a defensive aid, not a certification, and users are encouraged to review commands and rotate any exposed credential.

Where practical, processing happens locally in the browser. That design improves speed and reduces unnecessary collection of technical material. The project is not affiliated with Google, Anthropic, Microsoft, OpenAI, or the maintainers of any protocol mentioned unless a page explicitly says otherwise.

Editorial principles

• Explain checks and outputs in plain English.
• Keep examples useful, reproducible, and free of real credentials.
• Update material when the underlying specifications or common tools change.
• State important limitations instead of presenting automated output as professional advice.

Open source and corrections

The website source is available on GitHub. If you find an inaccurate rule, outdated example, accessibility problem, or broken link, please use the contact options below. Corrections that improve safety or technical accuracy are prioritized.