Summary
This is a bare-bones reference implementation from the Smithery cookbook, built to demonstrate the TypeScript SDK with their CLI tooling. It exposes a simple character counting tool and shows you how to structure a TypeScript MCP server using streamable HTTP transport instead of stdio. Think of it as a hello world with training wheels. The real value is in the source repo, which includes migration guides for moving existing stdio servers to HTTP, examples with session configuration, and both Python and TypeScript variants. You'd clone this to see how the pieces fit together before building your own server, not to actually use the character counter itself.
Keep your Mac awake while Claude Code and 40+ AI agents run. Sleeps when they're idle.
One time payment $9 →
Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →
Smithery Cookbook
The Smithery Cookbook provides code examples and guides designed to help developers build MCP (Model Context Protocol) servers and clients, offering copy-able code snippets that you can easily integrate into your own projects.
Prerequisites
To make the most of the examples in this cookbook, you'll need:
- Smithery CLI: Install with
npm install -g @smithery/clito access the interactive playground and development tools - Programming language runtimes for the examples you want to explore (Python 3.12+, Node.js 18+, etc.)
- Basic understanding of the Model Context Protocol (MCP)
Quick Start
-
Clone this repository:
git clone https://github.com/smithery-ai/smithery-cookbook.git cd smithery-cookbook -
Choose an example and follow its README:
cd servers/python/quickstart # Follow the README.md instructions -
Test with Smithery Playground:
npx @smithery/cli playground --port 8081Note: Replace
8081with the port your server is running on -
Deploy to Smithery (optional): Ready to share your MCP server? Deploy it here to host it on Smithery's platform.
Table of Recipes
Python Servers
- FastMCP Quickstart - Basic server with greeting tool
- FastMCP Advanced - Server with session configuration
- Migrate STDIO to HTTP - Server with custom Docker container
TypeScript Servers
- TypeScript Quickstart - Simple server with character counting tool
- TypeScript Session Config - Server with session configuration
- Migrate STDIO to HTTP - Smithery CLI - Server using Smithery CLI
- Migrate STDIO to HTTP - Custom Container - Server with custom Docker container
Development Workflow
Build and distribute your MCP servers with Smithery:
- Build your MCP server using the language and framework of your choice
- Test interactively with
npx @smithery/cli playground - Debug with real-time request/response inspection
- Deploy to Smithery's hosted platform - Deploy here
- Distribute your server gets its own page at
smithery.ai/server/{name}for others to discover and use
Explore Further
Looking for more resources to enhance your MCP development experience?
- Smithery Documentation - Complete guides and API reference
- Deploy Your MCP Server - Host your server on Smithery's platform
- Model Context Protocol Specification - Official MCP documentation
- Smithery Discord Community - Get help and share your projects
If you have ideas for new examples or guides, share them on the issues page.
License
MIT License
Featured
Keep your Mac awake while Claude Code and 40+ AI agents run. Sleeps when they're idle.
One time payment $9 →Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →