Custom-MCP-Server-for-AI-Agents

MCPioneer 🚀

Build Your Own Custom MCP Server for AI Agents

MCPioneer is a beginner-friendly project demonstrating how to create a custom MCP (Message/Command/Processing) server using Python. This server acts as a bridge to extend the capabilities of AI agents, allowing them to perform external tasks like calling APIs, accessing databases, and executing custom tools.

---

📚 What is MCP?

An MCP server provides tools, resources, and services that AI agents can access to enhance their functionalities. Think of it as a "toolbox" that an AI can open whenever it needs external help beyond its basic capabilities.

---

⚙️ Technologies Used

• Python 3.10+
• MXGp Python SDK
• UV Package Manager
• Claude Desktop (for testing)

---

🚀 Getting Started

1. Install UV (Better Package Manager)

• Windows:

  iwr -useb https://mirror.ghproxy.com/https://raw.githubusercontent.com/astral-sh/uv/main/scripts/install.ps1 | iex
  

• Mac/Linux:

  curl -LsSf https://astral.sh/uv/install.sh | sh
  

2. Initialize Project

uv init .
uv add mcp cli

3. Create a Simple MCP Server

Example main.py:

from mcp_servers.fast_mcp import FastMCP

server = FastMCP(name="DemoServer")

@server.tool
def add_numbers(a: int, b: int) -> int:
    return a + b

@server.resource
def greet(name: str) -> str:
    return f"Hello, {name}!"

server.run()

4. Install MCP Server into Claude Desktop

uv run mcp install main.py

5. Configure Claude Desktop

• Go to Settings → Developer → Edit Config.
• Add your server to the configurations manually if necessary.
• Restart Claude Desktop if you don't see the MCP server immediately.

---

🛠️ Project Structure

MCPioneer/
├── main.py
├── README.md
├── .venv/ (created automatically)
└── uv.toml

---

---

✨ Author

Created with ❤️ by Nithin.