MCP Boilerplate

# MCP Server Boilerplate (josharsh) **Repository**: https://github.com/josharsh/mcp-server-boilerplate ## Overview Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP. A solid, foundational starting point for MCP projects with emphasis on extensibility, security, and production readiness. ## Tech Stack - **Python** - Core language - **FastMCP** - MCP framework - **Pydantic** - Type-safe input validation - **Multiple Transports** - STDIO, SSE, HTTP support - **Docker** - Containerization support ## Project Structure ``` src/ ├── base/ # Core abstract base classes ├── tools/ # Modular tool implementations ├── resources/ # Static/dynamic resource management ├── prompts/ # Prompt generation and management ├── transports/ # Pluggable transport layer implementations ├── config.py # Environment and configuration handling └── server.py # Server instantiation logic ``` ## Key Features - ✅ Modular design with clear separation of concerns - ✅ Registry-based component registration - ✅ Multiple transport layers (STDIO, SSE, HTTP) - ✅ Type-safe input validation with Pydantic - ✅ Sandboxed file/directory operations - ✅ Comprehensive error handling - ✅ Production-ready logging and environment management - ✅ Docker containerization support ## Implementation Patterns - **Extensible Architecture**: Easy addition of tools, prompts, and resources - **Base Class Design**: Abstract base classes for core components - **Registry Pattern**: Centralized component registration - **Transport Abstraction**: Pluggable transport layer implementations - **Configuration Management**: Centralized environment handling - **Security Focus**: Input validation and sandboxed operations ## "Better Way" Characteristics - **Flexibility**: Multiple transport layer support - **Production Ready**: Comprehensive logging and error handling - **Security First**: Type validation and sandboxed operations - **Extensibility**: Easy to add new components - **Best Practices**: Clean architecture and separation of concerns ## Component Architecture - **Tools**: Modular tool implementations with base classes - **Resources**: Static and dynamic resource management - **Prompts**: Prompt generation and management system - **Transports**: Pluggable transport layer (STDIO/SSE/HTTP) - **Config**: Environment and configuration handling ## Unique Approaches - **Registry-Based Registration**: Centralized component management - **Transport Agnostic**: Supports multiple communication protocols - **Type Safety**: Pydantic validation throughout - **Sandboxed Operations**: Security-focused file operations - **Modular Design**: Clear separation enabling easy extension ## Recommended Use Cases - Production MCP server deployments - Projects requiring multiple transport protocols - Security-sensitive applications - Large-scale MCP implementations - Teams needing structured, extensible architecture ## Analysis This boilerplate excels in providing a production-ready, extensible foundation for MCP servers. The emphasis on modularity, type safety, and multiple transport support makes it ideal for complex projects requiring scalability and maintainability. The registry-based architecture and base class design patterns provide excellent foundations for large-scale MCP implementations.