Langgraph Coding Team

🚀 Collaborative Coding Robot Project

This is a collaborative coding robot that combines LangGraph and Model Context Protocol (MCP). The system consists of three parts: a gateway server, multiple MCP servers, and three collaborative Coding Agents.

🚀 Quick Start

1. Install Dependencies

# Install the agent package
pip install -e .

# Install the gateway package
cd gateway
pip install -e .
cd ..

2. Configure Environment Variables

Support multiple LLM providers through environment variables:

# LLM configuration - Support multiple providers:
LLM_MODEL=provider/model-name

# Supported providers and example models:
# - Anthropic: anthropic/claude-3-5-sonnet-20240620
# - OpenAI: openai/gpt-4
# - OpenRouter: openrouter/openai/gpt-4o-mini
# - Google: google/gemini-1.5-pro

# API Keys for different providers
OPENAI_API_KEY=your_openai_api_key
ANTHROPIC_API_KEY=your_anthropic_api_key
OPENROUTER_API_KEY=your_openrouter_api_key
GOOGLE_API_KEY=your_google_api_key

# OpenRouter configuration (if using OpenRouter)
OPENROUTER_BASE_URL=https://openrouter.ai/api/v1

3. Configure MCP Servers

The gateway server is configured through gateway/config.json. By default, two MCP servers are started:

{
  "mcp": {
    "servers": {
      "filesystem": {
        "command": "npx",
        "args": [
          "-y",
          "@modelcontextprotocol/server-filesystem",
          "/path/to/directory"
        ]
      },
      "memory": {
        "command": "npx",
        "args": [
          "-y",
          "@modelcontextprotocol/server-memory"
        ]
      }
    }
  }
}

You can add more servers in the official MCP server repository.

4. Start the Gateway Server

cd gateway
python -m mcp_gateway.server

The server will start on the default port 8808.

5. Configure the Agent

The connection between the agent and the gateway is configured in langgraph.json:

{
  "dependencies": ["."],
  "graphs": {
    "agent": "./src/react_agent/graph.py:graph"
  },
  "env": ".env",
  "mcp": {
    "gateway_url": "http://localhost:8808"
  }
}

6. Use the Agent

Open the folder in LangGraph Studio! The agent will automatically:

1. Connect to the gateway
2. Discover available tools
3. Start collaborative coding

✨ Features

Component Introduction

MCP Gateway Server

• Manage multiple MCP server instances.
• Provide a unified API interface for access tools.
• Handle communication with MCP servers.
• Expose tools through a simple HTTP interface.

MCP Servers

• Servers with specific functions:
  • GitHub Server: Support repository operations (read/write, list, search, create branches, create PRs, etc.).
  • More servers can be added to expand functionality.

Coding Agents

• Orchestrator: Collect context from human information and use MCP servers to access Linear and GitHub. Delegate to the Planner and Coder as needed.
• Planner: Receive requirements and code context, formulate plans, and propose multiple implementation suggestions. Does not use MCP.
• Coder: Receive code context and proposed implementations, and implement everything on different GitHub branches.

💻 Usage Examples

Available Tools

Filesystem Tools

• Read files
• Write files
• Traverse directories
• Delete files
• Copy files

Memory Tools

• Store context
• Manage sessions
• Persist data
• Query history

📚 Documentation

For more detailed information, please refer to:

• LangGraph Official Documentation
• MCP Protocol Documentation

🔧 Technical Details

Development Guide

Add a New Server

Add a new MCP server configuration in config.json and restart the gateway.

Customize the Agent

You can customize the agent's behavior by modifying graph.py and related components.

📄 License

This project uses the MIT license. For details, see the LICENSE file.