CodeBuddy MCP

An AI-powered development assistant that uses the Model Context Protocol (MCP) to connect LLMs with real-world developer tools. CodeBuddy can write and execute code, manage GitHub repositories, and deploy projects to Vercel and Render -- all through a conversational interface.

Features

• Multi-LLM Support -- Choose between Anthropic (Claude) or OpenRouter (GPT-4o, Gemini, and more)
• Sandboxed Code Execution -- Run Python and Bash safely inside Docker containers
• GitHub Integration -- Create repos, manage issues/PRs, search code, and push files via MCP
• Vercel Deployment -- Deploy projects, manage environment variables, and inspect deployments
• Render Deployment -- Deploy and manage cloud services on Render via MCP
• Multi-Turn Conversations -- Maintains full conversation history for context-aware assistance
• Workspace Management -- Timestamped session directories keep projects organized

Architecture

User Input
    |
    v
 main.py ──> API Provider Selection
    |
    v
 llm.py ──> LLM Communication (Anthropic / OpenRouter)
    |              |
    |         Tool Calls
    v              v
 mcp_connections.py ──> Tool Router
    |         |              |
    v         v              v
 GitHub    Render     docker_interpreter.py
  MCP       MCP              |
 Server    Server      Docker Container
                        (Python, Bash,
                         Vercel CLI)

Project Structure

File	Description
main.py	Entry point -- handles user input, API selection, and the interactive loop
llm.py	LLM handler for Anthropic and OpenRouter APIs with tool-calling support
mcp_connections.py	MCP connection manager -- routes tool calls to GitHub, Render, or Docker
docker_interpreter.py	Docker container manager for sandboxed code execution and file ops
custom_server.py	FastMCP server template for defining custom tools
Dockerfile	Container image with Python 3.12, Node.js 22, and Vercel CLI
test_docker.py	Verification script for testing the Docker setup

Prerequisites

• Python 3.12+
• Docker Desktop (running)
• Node.js 20+
• API keys for the services you want to use

Setup

1. Clone the repository

   git clone https://github.com/satwik6941/CodeBuddy-MCP.git
   cd CodeBuddy-MCP

2. Install dependencies

   pip install -r requirements.txt

3. Configure environment variables

   Create a .env file in the project root:

   GITHUB_PERSONAL_ACCESS_TOKEN=ghp_your_token_here
   CLAUDE_API_KEY=sk-ant-api03-your_key_here
   OPENROUTER_API_KEY=sk-or-v1-your_key_here     # Optional
   RENDER_API_KEY=rnd_your_key_here
   VERCEL_TOKEN=your_vercel_token_here

4. Build the Docker image

   docker build -t codebuddy-mcp .

5. Verify Docker setup (optional)

   python test_docker.py

Usage

python main.py

On launch, you will:

1. Describe your problem statement or task
2. Select an API provider (Anthropic or OpenRouter)
3. If OpenRouter, choose a model from the list or enter a custom one
4. Enter interactive mode where you can chat with the AI assistant

Type exit to quit the session.

Example Workflows

Create and deploy a web app:

"Build a simple Express API with a /health endpoint and deploy it to Vercel"

Analyze data:

"Read my CSV file, compute summary statistics, and generate a bar chart"

Fix a bug:

"Clone my repo, find why the login endpoint returns 500, and push a fix"

Available Tools

GitHub (via MCP)

Create/manage repositories, issues, pull requests, branches, and files. Search code across repos.

Docker (Code Execution)

Tool	Description
execute_python	Run Python code in a sandboxed container
execute_bash	Run Bash commands in the container
write_file	Create or write files in the workspace
read_file	Read file contents from the workspace
list_local_files	List files in the workspace directory
sync_files	Sync container files to local workspace

Vercel (Deployment)

Tool	Description
vercel_deploy	Deploy a project to Vercel
vercel_list_projects	List all Vercel projects
vercel_list_deployments	List deployments for a project
vercel_logs	View deployment logs
vercel_inspect	Inspect deployment details
vercel_env_add	Add environment variables

Render (via MCP)

Deploy services, manage environment variables, view logs, and scale services.

Custom Tools

Define your own tools in custom_server.py using the FastMCP framework.

Tech Stack

• Python 3.12 -- Core language
• MCP (Model Context Protocol) -- Tool integration framework
• Anthropic SDK -- Claude API access
• OpenAI SDK -- OpenRouter compatibility
• Docker -- Sandboxed code execution
• Vercel CLI -- Cloud deployment
• FastMCP -- Custom MCP server framework

License

This project is open source. See the repository for license details.