MCP server · Developer tools
Muninn code search server
by kamysh
Lets your AI search your codebase by meaning and by text, so it can find where things are implemented.
Muninn is a helper that gives your AI coding assistant a searchable index of your code. Instead of reading files one by one, the AI can ask Muninn where something lives and get back the relevant pieces. It is handy for developers working in large repositories, especially when they use several Git worktrees for different branches.
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 another tool. Muninn is that helper here: it connects your AI to a search index of your code. When you ask a question, the AI talks to Muninn, and Muninn looks things up in the index and hands back the matching code.
What this MCP server does
You ask your AI something like where a feature is implemented. The AI sends that question to Muninn through the MCP connection. Muninn searches its index of your repository using a mix of meaning-based and word-based search. It finds relevant source ranges and returns them to the AI. The AI then reads those pieces and explains or changes the code for you.
What you can do with it
- Find where a behavior is implemented in your code
- Search code by meaning, by exact words, or both together
- Follow recorded relationships between parts of the code
- Reuse the same index across linked Git worktrees
- Save and search notes about your repository
- List and delete saved notes
Try asking your AI
- “Where is the login validation handled in this repo?”
- “Show me the code that builds the user profile page.”
- “Search for anything related to payment retries and show the relevant files.”
- “What notes do we have about the database migration process?”
What it gives back to you
You get back source excerpts from your code that match your question, along with the file and location. For knowledge searches, you get your saved notes. The AI usually summarizes what it found and points you to the right place. Search results are evidence to inspect, not guaranteed answers, so the AI may read the code further to confirm.
Before you start
What you need
- PostgreSQL with pgvector, Apache AGE and pgcrypto (the postgres-ai project bundles these)
- The Muninn programs installed (CLI, index watcher and MCP server)
- An MCP client such as Claude Code or Codex
- An embedding backend, either local or a hosted service like Voyage or OpenAI
Good to know
Hosted embedding services send selected source code to the provider, and a remote database stores your index off your machine, so check your privacy settings before using them.
Install it with your AI
Add Muninn code search 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 Muninn code search 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 4,609 MCP servers here for you.
Who it's for
Developers and coding teams who work in large repositories and want their AI assistant to find code faster.
