boot
A starter template for building Model Context Protocol (MCP) applications with TypeScript
claude mcp add --transport stdio kirklin-boot-mcp npx -y boot-mcp
How to use
boot-mcp is a template for building MCP servers in TypeScript. It provides a structured setup for Resources (data access), Tools (actions and side-effectful operations), Prompts (LLM interaction templates), Roots (scopes), Transports (stdio and HTTP/SSE), and optional Sampling capabilities for agent-style workflows. This repository includes example servers that demonstrate how to expose data and functionality to LLMs across transports, enabling streamlined integration in CLI tools and web-based applications. Use the included examples to understand how to define resources, implement tools, and wire them to prompts and transports to create an end-to-end MCP experience.
Key capabilities you can leverage here include: creating text and binary Resources, implementing utility and data-processing Tools, crafting reusable Prompts for common workflows, managing file-system or URI-based Roots, and enabling transport-ready servers via Stdio or HTTP/SSE. The Quick Start section demonstrates running both a basic stdio server and an HTTP server, illustrating how MCP servers can operate in local CLI contexts or remote web environments. Explore the example servers and the structure under src and examples to see how each piece fits together and how to extend them for your own data or tools.
How to install
Prerequisites:
- Node.js (LTS version recommended) and npm/yarn/pnpm installed on your system
- Internet access to fetch dependencies
Installation steps:
- Clone the repository or open this project in your workspace.
- Install dependencies with your package manager of choice:
# npm
npm install
# pnpm
pnpm install
# yarn
yarn install
- Build or transpile if required by the project setup (refer to scripts in package.json). For many TypeScript MCP templates, the runtime can be executed directly after installation depending on the publishing setup.
- Start the server using the provided npm scripts (examples in the README). For instance:
# Start the stdio server (basic example)
pnpm run start:basic
# Start the HTTP server (SSE transport)
pnpm run start:http
- If you plan to publish or run via npx on a new environment, you can run directly with:
npx -y boot-mcp
Notes: Adjust Node.js version according to the template’s minimum requirements, and consult package.json for any environment-specific configuration or additional scripts.
Additional notes
Tips and caveats:
- This template supports multiple transports; choose stdio for local CLI usage or HTTP/SSE for remote integration.
- Review the src structure to understand where Resources, Tools, Prompts, Roots, and Transports are implemented and how they interact.
- If you customize environment-specific settings (e.g., ports, hostnames, authentication), consider adding them to an .env file or your deployment environment and loading them via process.env.
- Common issues often relate to TypeScript compilation steps or missing dependencies; ensure your npm install completes successfully and that your Node version aligns with the project’s requirements.
- When expanding functionality, keep the MCP interface consistent with the existing typing to ensure compatibility with LLMs and downstream clients.
Related MCP Servers
context7
Context7 MCP Server -- Up-to-date code documentation for LLMs and AI code editors
obsidian -tools
Add Obsidian integrations like semantic search and custom Templater prompts to Claude or any MCP client.
MiniMax -JS
Official MiniMax Model Context Protocol (MCP) JavaScript implementation that provides seamless integration with MiniMax's powerful AI capabilities including image generation, video generation, text-to-speech, and voice cloning APIs.
mcp-bundler
Is the MCP configuration too complicated? You can easily share your own simplified setup!
akyn-sdk
Turn any data source into an MCP server in 5 minutes. Build AI-agents-ready knowledge bases.
promptboard
The Shared Whiteboard for Your AI Agents via MCP. Paste screenshots, mark them up, and share with AI.