MCP Server
A production-grade Model Context Protocol server that enables secure management of context data through a React dashboard, supporting ZIP processing, context approval workflows, and system monitoring.
README
MCP Server - Production Monorepo
A production-grade Model Context Protocol (MCP) server with a beautiful React dashboard, built with security, scalability, and maintainability in mind.
🚀 Features
Backend (Node.js + Express + TypeScript)
- Full MCP Protocol Compliance - Complete implementation of the Model Context Protocol
- Secure ZIP Processing - Advanced security validation, zip bomb protection, path traversal prevention
- Context Management - Approval workflow for uploaded content
- System Snapshots - Comprehensive state snapshots with configurable options
- Production Security - Rate limiting, CORS, helmet, input validation
- Structured Logging - Winston-based logging with rotation and levels
- Health Monitoring - Comprehensive health checks and system statistics
Frontend (React + TypeScript + Tailwind)
- Modern Dashboard - Beautiful, responsive interface with dark/light mode
- Real-time Monitoring - Live system stats, logs, and activity feeds
- Secure File Upload - Drag-and-drop ZIP upload with progress tracking
- Context Browser - Search, filter, and manage uploaded contexts
- Log Viewer - Real-time log streaming with filtering and export
- Tool Management - MCP tool monitoring and configuration
MCP Tools
- load_zip - Securely extract and process ZIP archives
- mark_approved - Context approval workflow management
- snapshot - Create comprehensive system state snapshots
🏗️ Architecture
mcp-server/
├── backend/ # Node.js + Express API server
│ ├── src/
│ │ ├── mcp/ # MCP protocol implementation
│ │ ├── api/ # REST API routes and middleware
│ │ ├── utils/ # Utilities and security
│ │ └── types.ts # TypeScript definitions
├── dashboard/ # React frontend
│ ├── src/
│ │ ├── components/ # Reusable UI components
│ │ ├── pages/ # Application pages
│ │ └── api.ts # API client
├── data/ # Runtime data storage
│ ├── zips/ # Uploaded ZIP files
│ ├── extracted/ # Extracted file contents
│ ├── vds/ # Vector database storage
│ └── snapshots/ # System snapshots
└── logs/ # Application logs
🚦 Quick Start
Prerequisites
- Node.js 18+
- npm 8+
Local Development
- Clone and install dependencies:
git clone <repository>
cd mcp-server
npm install
- Start development servers:
npm run dev
This starts both the backend (port 3001) and frontend (port 5173) in development mode.
- Access the application:
- Dashboard: http://localhost:5173
- API: http://localhost:3001
- Health Check: http://localhost:3001/api/health
Production Build
npm run build
npm start
🔧 Configuration
Environment Variables
Create a .env file in the backend directory:
NODE_ENV=production
PORT=3001
CORS_ORIGIN=*
MAX_FILE_SIZE=52428800
LOG_LEVEL=info
RATE_LIMIT_ENABLED=true
Security Configuration
The server includes comprehensive security measures:
- File Upload Limits: 50MB max file size, ZIP files only
- Rate Limiting: 5 uploads per 15 minutes, 100 API calls per 15 minutes
- ZIP Security: Zip bomb protection, path traversal prevention, file type validation
- Input Validation: Joi-based validation for all endpoints
- CORS: Configurable cross-origin resource sharing
- Helmet: Security headers and CSP
📡 API Endpoints
MCP Tools
POST /api/mcp/tools/load-zip- Upload and extract ZIP filesPOST /api/mcp/tools/mark-approved- Approve/reject contextsPOST /api/mcp/tools/snapshot- Create system snapshots
Data Management
GET /api/context- List contexts with filtering and paginationGET /api/logs- Retrieve system logsGET /api/stats- System statistics and metricsGET /api/mcp/tools- List available MCP tools
System
GET /api/health- Health check endpoint
🔒 Security Features
ZIP File Processing
- File Type Validation: Only ZIP files accepted
- Size Limits: 50MB upload limit, 100MB extraction limit
- Zip Bomb Protection: Compression ratio analysis
- Path Traversal Prevention: Sanitized file paths
- Content Filtering: Allowed file extensions only
API Security
- Rate Limiting: Configurable request limits
- Input Validation: Comprehensive request validation
- Error Handling: Secure error responses
- CORS Configuration: Controlled cross-origin access
- Security Headers: Helmet.js integration
🚀 Deployment
Render (Recommended)
The project includes a render.yaml configuration for easy deployment:
- Connect your repository to Render
- The build and deployment will happen automatically
- Environment variables are configured in the YAML file
Manual Deployment
- Build the project:
npm run build
- Set environment variables:
export NODE_ENV=production
export PORT=3001
- Start the server:
npm start
📊 Monitoring
Dashboard Features
- System Overview: Uptime, file counts, memory/disk usage
- Real-time Logs: Live log streaming with filtering
- Context Management: Upload approval workflow
- Tool Monitoring: MCP tool usage statistics
Logging
- Structured Logging: JSON format with metadata
- Log Rotation: Automatic log file rotation
- Multiple Levels: Error, warn, info, debug
- Export Capability: Download logs for analysis
🧪 Testing
# Run backend tests
npm run test --workspace=backend
# Run frontend tests
npm run test --workspace=dashboard
# Run all tests
npm test
🤝 Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests for new functionality
- Submit a pull request
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
🆘 Support
For support and questions:
- Check the Issues page
- Review the API documentation
- Check the logs for error details
🔄 Version History
- v1.0.0 - Initial release with full MCP protocol support
- Production-ready security features
- Complete React dashboard
- Render deployment configuration
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。