CodeGraphContext MCP server

CodeGraphContext is a helper that reads the code on your computer and builds a map of how everything links together. Once it is connected to your AI assistant, you can ask questions about your project and get answers based on that map instead of guesswork. It is handy for developers and anyone who has to understand a codebase they did not write.

What this MCP server does

You point CodeGraphContext at a folder of code, and it reads through the files and builds a map of functions, classes, and how they call each other. That map is saved in a small database on your machine. When you ask your AI something about the code, the AI asks this helper, the helper looks in the map, and sends back the relevant pieces. You then see a plain answer in your chat, such as which function calls another one or where a class is used.

Click to zoom

What you can do with it

• Find who calls a particular function
• Trace a chain of calls across several files
• See the class hierarchy in your project
• Spot code that is never used
• Check which files import a module
• Watch a folder and update the map when files change
• Load a ready-made map for a well-known open source project

Try asking your AI

• “Which functions call the login method?”
• “Show me the chain of calls that starts at main.”
• “Is this helper function used anywhere in the project?”
• “What classes inherit from BaseHandler?”

What it gives back to you

You get answers in plain language in your chat, often with a short list of file names and function names. It can also show small diagrams or summaries of how things connect. If you ask it to index a folder, it tells you when the map is ready. Nothing in your code is changed; it only reads.

Before you start

Who it's for

Developers, maintainers, and anyone who needs to understand an unfamiliar codebase without reading every file.