MCP server · Developer tools

Metatron MCP server

Lets your coding AI look up your team's real coding conventions before it writes code.

Developer tools Coding For it and development Developer Tools

Metatron is a self-hosted helper that collects the unwritten rules of your codebase, like which patterns your team prefers and which approaches they already rejected. It runs on your own machine or server, and it serves those rules to your coding AI. It is handy for developers and teams who are tired of their AI assistant guessing at conventions.

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 else. This one connects your AI to Metatron, a local knowledge base of your project's decisions and conventions. When you ask your AI to write or change code, it can quietly check Metatron first, so it follows your team's way of doing things.

What this MCP server does

You ask your coding AI to work on a piece of your project. The AI calls this helper and asks what conventions apply there. Metatron looks through its stored decisions, which come from your git history and from feedback developers gave, and returns the most relevant ones. Your AI then writes code that matches those conventions instead of inventing its own. You can also have the AI report gaps it noticed, and those become new candidates for your team to review.

Flow diagram: you ask your AI “Check Metatron for our coding conventions here”, on your own computer the Metatron MCP server works with your computer, and you get back conventions and warnings.

Click to zoom

What you can do with it

Look up your team's preferred patterns before writing code
See approaches your team already tried and rejected
Find edge cases and gotchas recorded for a part of the codebase
Report gaps or unclear conventions back as feedback
Review and approve new candidate decisions before they count
Keep decisions in sync between plain files and the server

Try asking your AI

“Before you change the payment module, check Metatron for our conventions there.”
“What pattern does our team use for error handling in this service?”
“I noticed the docs never say how we name migration files. Submit that as feedback to Metatron.”
“Show me the decisions Metatron has about our API layer.”

What it gives back to you

You get back short answers in the chat: the relevant conventions, the reasoning behind them, and any warnings about past mistakes. It can also return lists of decisions or candidates when you ask for an overview. When your AI submits feedback, Metatron stores it and later turns it into new candidate decisions for your team to review. Nothing becomes an official rule without a human approving it.

Before you start

What you need

Git installed on your computer
Python 3.12 or newer, or Docker
An Anthropic API key for the steps that read your git history (the serving part needs no key)
A code repository you want Metatron to learn from

Good to know

Metatron reads your git history and stores your team's internal conventions, so treat the database and files as private, and remember that extraction steps send structural signals to Anthropic's model.

Install it with your AI

Add Metatron 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 Metatron 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.