ERPNext MCP Server

ERPNext MCP Server

Enables ERPNext management, file operations, read-only database access, and ERPNext API integration through a standardized MCP server.

Category
访问服务器

README

Standardized MCP Development on Frappe/ERPNext

ERPNext MCP Server

ERPNext management, file operations, read-only database access, and ERPNext API integration

The Development to Production Workflow

This guide establishes a standardized approach for developing Model Context Protocol (MCP) servers on Frappe/ERPNext, inspired by the Doppio SPA development workflow.

Development Mode (Port 8080)

  • Live development environment with hot reloading
  • Direct debugging and testing
  • Clear separation from production

Production Mode (Port 8100)

  • Integrated with NGINX and supervisor
  • Secure and scalable
  • Follows Frappe's established patterns

Setting Up the Development Environment

1. Install the Framework

Add the MCP development framework to your app's hooks.py:

# In your app's hooks.py
commands = [
    "your_app.mcp_dev.commands"
]

2. Initialize the Development Environment

# Initialize development environment
bench --site your-site.com mcp-config --app your_app_name

# Configure development port (default: 8080)
bench --site your-site.com mcp-config --dev-port 8080

# Configure production port (default: 8100)
bench --site your-site.com mcp-config --prod-port 8100

3. Start Development Server

# Start development server
bench --site your-site.com mcp-dev

This will:

  • Start your MCP server on port 8080
  • Watch for file changes and auto-reload
  • Stream logs to the console

Development to Production Workflow

1. Develop on Port 8080

During development:

  • Your MCP server runs on port 8080
  • Changes to Python files trigger auto-reload
  • You can debug and test in real-time

2. Build for Production

When ready to deploy:

# Build for production
bench --site your-site.com mcp-build

This will:

  • Configure NGINX to serve your MCP server on port 8100 under /mcp/
  • Set up supervisor to keep your MCP server running
  • Prepare all necessary configuration files

3. Deploy to Production

# Deploy to production
bench --site your-site.com mcp-deploy --reload-nginx --restart-supervisor

This will:

  • Apply the NGINX configuration
  • Start the MCP server under supervisor
  • Make it available at https://your-site.com/mcp/

Configuration Management

The framework provides a configuration system to manage your MCP development environment:

# View current configuration
bench --site your-site.com mcp-config

# Update entry point
bench --site your-site.com mcp-config --entry-point "your_app.mcp_server:server"

# Add watched paths
bench --site your-site.com mcp-config --add-watch "apps/your_app/your_app/mcp/*.py"

# Set environment variables
bench --site your-site.com mcp-config --set-env "DEBUG" "true"

Benefits of the Standardized Approach

  1. Development-Production Parity:

    • Same code runs in both environments
    • No surprises when deploying
  2. Developer Experience:

    • Hot reloading during development
    • Clear separation of concerns
    • Unified configuration management
  3. Production Robustness:

    • Proper integration with NGINX
    • Supervisor manages process lifecycle
    • Standard Frappe deployment patterns
  4. Team Collaboration:

    • Consistent development approach
    • Standard commands and workflows
    • Easy onboarding for new developers

Best Practices

1. Structuring Your MCP Server

Organize your MCP server code with a clean separation of concerns:

your_app/
├── your_app/
│   ├── mcp_server.py       # Main server entry point
│   ├── mcp/
│   │   ├── __init__.py     # Package initialization
│   │   ├── tools/          # MCP tools implementation
│   │   ├── resources/      # MCP resources implementation
│   │   └── prompts/        # MCP prompts implementation
│   └── mcp_dev.py          # Development framework

2. Version Control

  • Include the development configuration in version control
  • Exclude environment-specific settings
  • Document required environment variables

3. Testing

  • Create specific tests for MCP functionality
  • Use both unit tests and integration tests
  • Test in both development and production modes

4. Documentation

  • Document your MCP server's capabilities
  • Create examples of using your tools and resources
  • Include setup instructions for both development and production

Advanced Topics

1. Multiple MCP Servers

For larger applications, you might need multiple MCP servers:

# Configure multiple entry points
bench --site your-site.com mcp-config --entry-point "your_app.mcp_server_1:server"
bench --site your-site.com mcp-config --dev-port 8081

# Start specific server
bench --site your-site.com mcp-dev --port 8081

2. Security Considerations

  • Use proper authentication in production
  • Consider rate limiting for public-facing servers
  • Implement proper error handling and logging

3. Performance Optimization

  • Profile your MCP tools for performance bottlenecks
  • Consider caching frequently accessed data
  • Use asynchronous operations for I/O-bound tasks

By following this standardized approach, you can develop MCP servers for Frappe/ERPNext that are robust, maintainable, and follow established patterns familiar to Frappe developers.

License

mit

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选