MCP Basics

This repo is dedicated to learning how to create MCPs and use them with local solutions.

📚 Each MCP created has a dedicated tutorial: check them out here.

The first iteration of this project is to create an in-depth guide into what MCP is and how to use it. We'll go through an example of using the LangChain's docs MCP server in LM Studio.

The next iteration is to create a step-by-step guide into how to create our own MCP servers. We'll see how to create some simple web crawling servers using Crawl4AI, then test them out in LM Studio.

After that, I plan to create a step-by-step guide demonstrating how to use both Crawl4AI and Milvus MCP servers to gather, store, and query research from the web with the agentic platform OpenCode.

Finally, additional modular features that enhance deep research can be added to the OpenCode setup.

Getting started

1. To get started with the repo, clone it, create a virtual environment, and install all the necessary packages

git clone https://github.com/anima-kit/mcp-basics.git
uv venv venv
venv/Scripts/activate
uv pip install -r requirements.txt

For the Crawl4AI MCP, you'll also need to run:

playwright install

2. To run any of the MCP servers:

fastmcp dev mcp-server.py

Examples

• Calculator (mcp-calculator.py): A very simple calculator MCP that outlines how to use various aspects of the FastMCP library
• Adaptive web crawler (mcp-crawl4ai.py): A minimum web crawler MCP using the adaptive method from the Crawl4AI library.

Tech

To create web crawling tools

• FastMCP: To create MCP servers

Contributing

See the contributing guide

License

MIT: See the license