Catalogue of Tools & Metrics for Trustworthy AI

These tools and metrics are designed to help AI actors develop and use trustworthy AI systems and applications that respect human rights and are fair, transparent, explainable, robust, secure and safe.

TrustedMCP



TrustedMCP

TrustedMCP is an online directory that gives security grades to Model Context Protocol (MCP) servers. MCP servers are the connectors that let AI agents use external tools, data and services. Most MCP directories simply list servers. TrustedMCP grades each one from A to D, so users can see its security weaknesses before connecting it to an agent. It currently covers 158 servers in categories such as developer tools, databases, payments, cloud hosting and collaboration. It is run by a small team that also builds custom MCP servers.

Each server is assessed on four checks using public information only:

  • Authentication: whether scoped, revocable sign-in is the default.
  • Maintenance: how recently and by how many people the server is maintained.
  • Known vulnerabilities: whether published vulnerabilities are patched promptly.
  • Transport safety: whether connections use encryption and avoid running system commands on input controlled by the model.

Authentication carries the most weight. Servers that reach a user’s account on another platform by an unofficial route are capped at grade C, because the platform may block the account.

Listings show the grade, the result of each check and the first thing to fix, alongside community reviews. Grades are re-checked after major releases, newly disclosed vulnerabilities, or every quarter. The grades are also available through a free read-only application programming interface (API), and through a TrustedMCP MCP server that checks the servers in a user’s own configuration. The grades are not penetration tests or code audits.

Use Cases

There is no use cases for this tool yet.

Would you like to submit a use case for this tool?

If you have used this tool, we would love to know more about your experience.

Add use case
Partnership on AI

Disclaimer: The tools and metrics featured herein are solely those of the originating authors and are not vetted or endorsed by the OECD or its member countries. The Organisation cannot be held responsible for possible issues resulting from the posting of links to third parties' tools and metrics on this catalogue. More on the methodology can be found at https://oecd.ai/catalogue/faq.