mcp-agents

# MCP AI Development Team A sophisticated AI development team built with Model Context Protocol (MCP) for GitHub automation. Deploy anywhere with Docker or cloud platforms like Render. ## 🚀 Features - **Multi-Role AI Agents**: 5 specialized roles (Backend, Frontend, QA, DevOps, Security) - **Smart Role Selection**: Automatically chooses appropriate expertise based on file changes - **GitHub Integration**: Webhooks for push events, PR analysis, and automated improvements - **PostgreSQL Logging**: Complete audit trail of all AI actions - **Cloud-Ready**: Deploy to Render, Railway, or any cloud platform ## 🏗️ Architecture ``` GitHub Events → Webhook Server → AI Agent → GitHub Actions ↓ PostgreSQL Database Logging ``` ## 🤖 AI Roles - **Backend Engineer**: Code review, performance, security analysis - **Frontend Engineer**: UI/UX, accessibility, component architecture - **QA Engineer**: Test plans, edge cases, quality assurance - **DevOps Engineer**: Infrastructure, CI/CD, deployment optimization - **Security Engineer**: Vulnerability analysis, secure coding practices ## 🔧 Environment Variables ```bash # Required ANTHROPIC_API_KEY=your_anthropic_key_here GITHUB_TOKEN=your_github_token_here # Optional GITHUB_WEBHOOK_SECRET=optional_webhook_secret DATABASE_URL=postgresql://user:pass@host:port/db PORT=8000 HOST=0.0.0.0 ``` ## 🚀 Deployment ### Render (Recommended) 1. Fork this repository 2. Connect to Render 3. Set environment variables 4. Deploy as Web Service ### Docker ```bash # Build and run locally docker build -t mcp-agents . docker run -p 8000:8000 --env-file .env mcp-agents ``` ### Local Development ```bash # Install dependencies pip install -r requirements.txt # Set environment variables export ANTHROPIC_API_KEY="your_key" export GITHUB_TOKEN="your_token" # Run the server python webhook_server.py ``` ## 📡 API Endpoints - **GET /** - Health check - **POST /webhook/github** - GitHub webhook receiver - **GET /status** - Service status and recent activity - **POST /webhook/manual** - Manual testing endpoint - **GET /runs/{run_id}** - Detailed run information ## 🔧 GitHub Webhook Setup 1. Go to your repository **Settings → Webhooks** 2. Add webhook with: - **URL**: `https://your-deployment-url.com/webhook/github` - **Content type**: `application/json` - **Events**: Push, Pull requests - **Secret**: Your webhook secret (optional) ## 🗄️ Database Schema The system requires PostgreSQL with these tables: - `agent_runs` - Track AI executions - `agent_actions` - Log GitHub actions - `agent_memory` - Store context and preferences See `init.sql` for complete schema. ## 🧪 Testing ```bash # Health check curl https://your-deployment-url.com/ # Manual test curl -X POST https://your-deployment-url.com/webhook/manual \ -H "Content-Type: application/json" \ -d '{ "event_type": "push", "repository": {"full_name": "youruser/yourrepo"}, "commits": [{"modified": ["src/app.py"], "message": "Fix bug"}] }' ``` ## 📊 Monitoring View recent activity and stats: ```bash curl https://your-deployment-url.com/status ``` ## 🔮 Future Enhancements - Real MCP integration with official GitHub MCP server - Vector memory for code context and learning - Multi-repository analysis - Slack/Discord notifications - Human approval workflows ## 📝 License MIT License - see LICENSE file for details