code-buddy
Your AI-powered coding companion for Claude Desktop. MCP server with 23+ dev tools for file operations, git integration, code analysis, and more. Works with any project on your system.
claude mcp add --transport stdio abhi-vish-code-buddy uv --directory C:\path\to\code-buddy run python -m src.server.main \ --env PROJECT_ROOT="C:\path\to\code-buddy" \ --env ALLOW_EXTERNAL_PATHS="true"
How to use
Code Buddy is an MCP server that provides an AI-powered coding assistant with a rich set of tools for working with files, directories, code analysis, Git, and command execution. Once configured in Claude Desktop or your MCP client, you can ask the agent to read, edit, and manage code and project files, analyze code structure, format and lint Python files with Black and Ruff, run shell or Python commands, and interact with Git repositories. The included tools are organized into categories such as File Tools, Directory Tools, Search Tools, Code Tools, Git Tools, and Command Tools, enabling you to perform complex project operations through natural language prompts. With real-time streaming responses and support for custom working directories, Code Buddy aims to streamline common development workflows from within Claude Desktop or another MCP client.
How to install
Prerequisites:
- Python 3.13 or higher
- uv package manager (uv project) installed and available in PATH
- Claude Desktop installed for MCP integration (optional, if using Claude Desktop)
Setup steps:
- Clone the repository and navigate into the project directory:
git clone https://github.com/Abhi-vish/code-buddy.git
cd code-buddy
- Install and configure uv (or install project dependencies as specified in the repo):
# If your system uses uv as shown in the README
uv sync
- Prepare your OpenAI API key (for the agent functionality requiring OpenAI):
# Create a .env file in the project root
echo "OPENAI_API_KEY=your-api-key-here" > .env
- Run or test the MCP server using uv as configured (the example shown for Claude Desktop):
# Depending on how you launch, you might run the server module directly
uv run python -m src.server.main
- Ensure Claude Desktop or your MCP client is configured to point at the server using the provided mcpServers configuration (see the example in the README).
Additional notes
Tips and common considerations:
- Ensure OPENAI_API_KEY is set in environment or .env for any OpenAI-powered features.
- If you enable ALLOW_EXTERNAL_PATHS, be mindful of security and only point to trusted directories.
- Maintain the correct cwd/directory paths when configuring in Claude Desktop to ensure the MCP server can access project files.
- The server supports absolute and relative paths; use absolute paths when configuring external projects to avoid working directory confusion.
- If commands hang or time out, adjust timeouts in your MCP client and verify that required tools (uv, Python) are on PATH.
- Update and align your Python dependencies as the project evolves, especially with Black, Ruff, and code analysis utilities.
Related MCP Servers
mcp -tree-sitter
MCP Server for Tree-sitter
task-orchestrator
A light touch MCP task orchestration server for AI agents. Persistent work tracking and context storage across sessions and agents. Defines planning floors through composable notes with optional gating transitions. Coordinates multi-agent execution without prescribing how agents do their work.
fast-filesystem
A high-performance Model Context Protocol (MCP) server that provides secure filesystem access for Claude and other AI assistants.
memcord
🧠 Privacy-first MCP server for AI memory management. Save, search & organize chat history with intelligent summarization.
slack
Session-based Slack MCP for Claude and MCP clients: local-first workflows, secure-default HTTP.
openapi -swagger
Solve AI context window limits for API docs | Convert any Swagger/OpenAPI to searchable MCP server | AI-powered endpoint discovery & code generation | Works with Cursor, Claude, VS Code