add
A minimal MCP server example project to help understand how the MCP system operates.
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: add. It is used for: A minimal MCP server example project to help understand how the MCP system operates. Full Skill content: https://321skill.com/skills/add/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
This Skill addresses the difficulty developers face when first encountering the Model Context Protocol (MCP) in quickly setting up and running a minimal example project. In practical development, developers often need to understand the interaction flow between MCP servers and clients, but official documentation may not be intuitive enough, resulting in a steep learning curve.
Usage is straightforward: you only need to clone the repository, create a virtual environment, install dependencies, and add your API key to the .env file. Afterwards, you can configure different MCP servers by editing the server.json file and run client.py to input user queries for testing. The entire process clearly demonstrates the standard structure and operation of an MCP project.
It is well-suited for developers who need to quickly get started with the MCP protocol, conduct proof-of-concept work, or create teaching demonstrations. This is especially true for full-stack or backend developers already familiar with Python environments who wish to integrate LLM services (like DeepSeek) into their own AI applications via the MCP protocol.
It is recommended to use this tool uniformly when learning the MCP protocol or conducting functional demonstrations. Note that it is primarily an educational and demonstration project, providing a basic framework and file structure. However, specific server functionalities (e.g., tasklist.py, weather.py) require developers to extend and implement them according to their own needs.
Key Features
Unlike the complex official MCP examples, it provides a minimal, "Hello World"-level project with a clear file structure. This allows developers to immediately grasp the core components and operational flow of MCP without being overwhelmed by various advanced features.
Limitations
This project depends on a specific Python environment (e.g., the `uv` tool) and an OpenAI-compatible LLM API. Additionally, the server functionalities in the example are relatively basic and require self-extension to implement complex business logic.
FAQ
Which LLM services does this project support?
The project is configured by default for the DeepSeek API. However, in theory, any LLM service providing an OpenAI-compatible API (such as OpenAI, Azure OpenAI, etc.) can be used by modifying the `OPENAI_BASE_URL` and `OPENAI_MODEL` in the `.env` file.
How do I add my own MCP server?
In the `servers` array within the `server.json` file, add a new server configuration following the existing format, specifying the `command` and `args` to point to your server script.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/add/raw/index.md to read the original Skill definition (Markdown format) for add, and install it according to the instructions.
Raw Markdown URL for AI: /skills/add/raw/index.md