mcp-agent-orchestrator

Multi-Agent Development System

A sophisticated system that coordinates multiple AI coding assistants (Cursor and Claude Code) acting as specialized developers. Uses Model Context Protocol (MCP) for tool integration and custom coordination layers for agent communication and workflow orchestration.

🚀 Overview

This project implements a multi-agent development team where AI coding assistants collaborate like human developers. Each agent specializes in different roles (BA/PM, Backend Dev, Frontend Dev, Testers, DevOps) and communicates through MCP servers.

Key Features

• Platform-Agnostic Architecture: Works with any MCP-compatible AI coding assistant
• Distributed Coordination: Event-driven messaging between agents
• Specialized Roles: Each agent has specific capabilities and responsibilities
• Fault Tolerance: Built-in circuit breakers and rollback mechanisms
• Observable: Comprehensive monitoring and tracing

📋 Documentation

• System Design Document - Complete architecture and design
• Implementation Backlog - Phased development plan
• Technical Feasibility Assessment - Detailed analysis
• Analysis Summary - Executive summary

🏗️ Architecture

Core Components

1. MCP Servers - Protocol implementation for agent coordination

   • Task Coordinator
   • Message Queue
   • Shared Workspace
   • Documentation Hub

2. Agent Framework - Base classes and utilities for AI agents

   • Base Agent Class
   • MCP Client
   • State Manager
   • Communication Layer

3. Specialized Agents

   • Business Analyst/PM (Orchestrator)
   • Backend Developer
   • Frontend Developer
   • Various Testing Agents
   • DevOps Agent

🛠️ Technology Stack

• MCP Servers: Node.js/Python
• Agent Framework: Python with async/await
• Message Queue: Redis/RabbitMQ
• State Management: Event sourcing
• Infrastructure: Docker, Kubernetes

🚦 Getting Started

Prerequisites

• Python 3.9+
• Node.js 18+ LTS
• Docker and Docker Compose
• Git

Installation

1. Clone the repository:

git clone https://github.com/yourusername/multi-agent-dev-system.git
cd multi-agent-dev-system

2. Set up Python environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements.txt

3. Install Node.js dependencies:

npm install

4. Start Docker services:

docker-compose up -d

🧪 Development

Test-Driven Development

This project follows strict TDD practices:

• Write tests first (Red)
• Implement minimum code (Green)
• Refactor (Refactor)
• Minimum 85% coverage required

Running Tests

# Python tests
pytest --cov=agent-framework --cov-report=html

# Node.js tests
npm test

# All tests
make test-all

Code Quality

• Python: Black formatter, PEP 8 compliance
• JavaScript: ESLint with Prettier
• Pre-commit hooks for consistency

📚 Project Structure

.
├── mcp-servers/            # MCP server implementations
├── agent-framework/        # Core agent framework
├── configs/               # Configuration files
├── docs/                  # Documentation
├── scripts/               # Utility scripts
├── tests/                 # Test suites
├── docker-compose.yml     # Local development services
└── .cursorrules           # Development guidelines

🚀 Roadmap

Phase 0: Foundation (Current)

• Basic MCP server template
• Multi-instance coordination POC
• Simple message passing

Phase 1: Core Infrastructure

• Task Coordinator implementation
• Message Queue server
• Base agent framework

Phase 2: Essential Agents

• BA/PM Agent
• Backend Developer Agent
• Frontend Developer Agent

See Implementation Backlog for detailed roadmap.

🤝 Contributing

We welcome contributions! Please read our contributing guidelines before submitting PRs.

Development Process

1. Fork the repository
2. Create a feature branch (feature/amazing-feature)
3. Commit changes with conventional commits
4. Ensure tests pass with >85% coverage
5. Submit a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Model Context Protocol - The foundation for agent communication
• Anthropic - For Claude and the MCP specification
• Cursor - AI-powered code editor
• Claude Code - AI coding assistant

📞 Contact

• GitHub Issues: For bug reports and feature requests
• Discussions: For questions and community support

---

Note: This project is in active development. The architecture and APIs may change as we iterate on the design.