LangGraph Coding Agent Team
Createddanmas0n
Creates multiple implementations of proposed software features using coding agents and a unified API for tool access.
About
This project implements a team of coding agents utilizing LangGraph and the Model Context Protocol (MCP). The agents collaborate to take requirements and code context, generating multiple implementations of proposed features, allowing human operators to choose their preferred approach. It originated from the Anthropic MCP Hackathon and has evolved into a standalone project, showcasing the power of agent collaboration in software development.
Key Features
- Supports multiple LLM providers (Anthropic, OpenAI, OpenRouter, Google)
- Utilizes LangGraph for agent orchestration
- Employs multiple coding agents for parallel implementation
- Integrates with GitHub for code management (branching, PR creation)
- 18 GitHub stars
- Leverages Model Context Protocol (MCP) for tool access
Use Cases
- Collaborative coding between agents
- Rapid prototyping of software features
- Automated feature implementation with multiple options