MCP server · Security

Attestable MCP server

Lets your AI check that the server it talks to is running the exact code it claims to run.

This is a small server that your AI can connect to, with one extra trick: it can prove to your AI which code it is actually running. It is made for people who care about trust, like security teams, developers, and anyone handling sensitive data. If you just want your AI to answer questions, this is not the tool for you.

What is an MCP server? The 30-second version

On its own, your AI can only chat with you. An MCP server is a small helper program that gives your AI a new skill or a connection to something outside the chat. This particular helper runs inside a protected area of a computer called a trusted execution environment, and it hands your AI a kind of digital certificate that says: this is the exact code I am running. Your AI can check that certificate before it trusts anything the server says back.

What this MCP server does

You point your AI at this server. Before connecting, the server sends a certificate during the normal secure handshake that every website uses. That certificate contains proof of the code running inside a protected part of the machine, signed in a way your AI can verify. If the proof matches what you expect, your AI connects and uses the server. If it does not match, your AI can refuse to connect.

Flow diagram: you ask your AI “Is this server safe to trust?”, the Attestable MCP server connects it to Attestable MCP server, and you get back yes or no, with the details.

Click to zoom

What you can do with it

Verify that a server is running the exact code it claims to run
Check the certificate a server sends before trusting it
Confirm the code matches a build published on GitHub
Run the same check on your own machine without special hardware
Run the check on real secure hardware for stronger proof

Try asking your AI

“Check the certificate this server sends and tell me if the code matches the published build”
“Connect to the attestable server and confirm it is running the expected code before we continue”
“Show me what proof this server provides about the code it is running”
“Is this server safe to trust, based on the certificate it just sent?”

What it gives back to you

You get a yes or no answer about whether the server's proof checks out, plus the details behind it, like the code measurements and the certificate chain. In the chat, this usually looks like a short summary with the key values listed. If something does not match, you will see that too, so you know not to trust the connection.

Before you start

What you need

A computer or server that can run Docker
For real secure hardware: an Intel machine with SGX support
For testing only: a regular Linux machine, no special hardware needed

Good to know

This is a demonstration project, so treat it as a learning example rather than something to put in front of real users without a security review.

Install it with your AI

Add Attestable MCP server to your AI, no technical skills needed

You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.

Sign in to get the install prompt

Members get a ready-made prompt that lets the Claude desktop app check Attestable MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.