CodeMind

AI-powered code understanding for modern codebases.

CodeMind is a VS Code extension that helps developers understand, explore, and reason about large codebases using AI. Instead of relying on generic AI knowledge, CodeMind indexes your project locally, builds a semantic representation of your source code, and answers questions using the actual code in your repository.

✨ Features

• 🧠 AI-powered code understanding
• 📂 One-click workspace indexing
• 🔍 Semantic code search
• 💬 Ask questions about your codebase in natural language
• ⚡ Automatic re-indexing when files change
• 🏗️ Tree-sitter based parsing for accurate code analysis
• 🔒 Local-first architecture (your source code stays on your machine)

🚀 Supported Languages

Current version supports:

• Python
• JavaScript
• TypeScript
• React (JSX / TSX)
• Java

More languages will be added in future releases.

How CodeMind Works

1. Open your project in VS Code.
2. Index the workspace.
3. CodeMind parses your source code using Tree-sitter.
4. Code chunks are embedded and stored locally.
5. Ask questions in natural language.
6. CodeMind retrieves relevant context and generates accurate answers using AI.

Your source code is never uploaded as a complete repository.

Installation

Step 1 — Install the VS Code Extension

Install CodeMind from the VS Code Marketplace.

Step 2 — Install the Backend

CodeMind requires a local Python backend.

Create a virtual environment (Recommended)

Windows (PowerShell)

python -m venv .venv
.\.venv\Scripts\Activate.ps1

Windows (CMD)

python -m venv .venv
.\.venv\Scripts\activate.bat

macOS / Linux

python3 -m venv .venv
source .venv/bin/activate

Install the backend

pip install codemind-ai

Start the backend

codemind

You should see:

CodeMind backend running on http://127.0.0.1:8000

Keep this terminal running while using the extension.

Getting Started

Open a project folder in VS Code.

Run:

CodeMind: Index Workspace

Wait for indexing to complete.

Then run:

CodeMind: Ask Question

Example questions:

• Explain this project architecture.
• Where is user authentication implemented?
• How does the login flow work?
• Which files call this function?
• Explain this controller.
• Show me database interactions.
• Find all API endpoints.
• Explain this class.

Commands

Privacy

CodeMind follows a local-first architecture.

• ✅ Source code is indexed locally.
• ✅ Vector database is stored locally.
• ✅ Project files remain on your machine.
• ✅ Only the retrieved context required for answering your question is sent to the configured AI model.

Requirements

• VS Code 1.90+
• Python 3.10 or later
• Internet connection for AI responses

Troubleshooting

Backend not running

Create a virtual environment (recommended):

python -m venv .venv

Activate it:

Windows PowerShell

.\.venv\Scripts\Activate.ps1

Install backend:

pip install codemind-ai

Start backend:

codemind

"Could not connect to backend"

Verify that the backend is running.

Open:

http://127.0.0.1:8000/health

If the backend is running, it should return a health response.

Backend command not found

If codemind is not recognized:

• Ensure Python is installed.
• Ensure the Python Scripts directory is included in your system PATH.
• Restart your terminal after installation.

Roadmap

Upcoming features:

• Call Graph visualization
• Control Flow Graph generation
• Architecture diagrams
• Dependency Graph
• Cross-file workflow tracing
• Incremental indexing
• More programming language support
• Local LLM support
• Graph database integration
• Code impact analysis

Contributing

Contributions, feature requests, and bug reports are welcome.

Please open an issue or submit a pull request on GitHub.

License

MIT License

Author

Anubhav Pandey Ayush Sharma

Software Engineer passionate about AI-powered developer tools, static code analysis, and intelligent software engineering.

⭐ If you find CodeMind useful, consider giving the project a star on GitHub!