MCP Server Demo

Prerequisites

• uv - An extremely fast Python package and project manager, written in Rust.

Overview

This guide walks through creating a simple Model Context Protocol (MCP) server and testing it with both the MCP Inspector and Claude Desktop. The content is based on the modelcontextprotocol/python-sdk README, with additional notes for clarity.

Steps to Create an MCP Server

1. Environment Setup

uv init mcp-server-demo
cd mcp-server-demo

# Specify Python version
echo "3.12" > .python-version

# Install Python before creating a new virtual environment
uv venv 

# Activate virtual environment
source .venv/bin/activate.fish

# Verify virtual environment (check if using the correct Python environment)
which python
path/to/mcp-server-demo/.venv/bin/python

# Install MCP dependencies (added to pyproject.toml)
uv add "mcp[cli]"

# Verify MCP dependencies were installed correctly
uv tree

Resolved 28 packages in 3ms
mcp-server-demo v0.1.0
└── mcp[cli] v1.6.0
    ├── ...

2. Creating a Simple MCP Server

# server.py
from mcp.server.fastmcp import FastMCP

# Create an MCP server
mcp = FastMCP("Demo")

# Add an addition tool
@mcp.tool()
def add(a: int, b: int) -> int:
    """Add two numbers"""
    return a + b


# Add a dynamic greeting resource
@mcp.resource("greeting://{name}")
def get_greeting(name: str) -> str:
    """Get a personalized greeting"""
    return f"Hello, {name}!"

3. Testing with MCP Inspector

Launch MCP Inspector

uv run mcp dev server.py

Starting MCP inspector...
⚙️ Proxy server listening on port 6277

Testing the add Tool

1. Open the MCP Inspector
2. Click "Connect"
3. Click the "Tools" tab
4. Click "List Tools" inside the "Tools" tab

Testing the get_greeting Resource

1. Open the MCP Inspector
2. Click "Connect"
3. Click "Resources" Tab
4. Click "List Templates"
5. Find "get_greeting"
6. Type a name (e.g., "John")
7. Click "Read Resources"

Example response:

{
  "contents": [
    {
      "uri": "greeting://John",
      "mimeType": "text/plain",
      "text": "Hello, John!"
    }
  ]
}

4. Testing with Claude Desktop

Install the MCP Server in Claude Desktop

uv run mcp install server.py

[/dd/MM/YY HH:MM:SS] INFO     Added server 'Demo' to Claude config            claude.py:129
                     INFO     Successfully installed Demo in Claude app          cli.py:467

Verify Configuration

# For macOS/Linux
cat ~/Library/Application\ Support/Claude/claude_desktop_config.json

Example configuration:

{
    "mcpServers": {
        "Demo": {
            "command": "/absolute/path/to/uv",
            "args": [
                "run",
                "--with",
                "mcp[cli]",
                "mcp",
                "run",
                "/absolute/path/to/mcp-server-demo/server.py"
            ]
        }
    }
}

Note: After executing uv run mcp install server.py, the command is set as just "uv" in the configuration, but it requires an absolute path to work properly.

Using Claude Desktop with MCP

Checking Server Status

The Claude.app interface provides basic server status information:

1. Click the connect icon to view:
   • Connected servers
   • Available prompts and resources
2. Click the hammer icon to view:
   • Tools made available to the mode

• Debugging in Claude Desktop

Type in Claude Desktop:

Please use the add tool to sum 10 and 20

Using the get_greeting Resource

Note: There appears to be an issue retrieving resources from Claude Desktop. This section needs further investigation.

FAQ

How does Claude Desktop select which MCP server tools to use?

• Claude connects to all configured MCP servers but selects tools based on the content of the user's query
• If the question content is not suitable for using tools, the tools will not be used

References

• Model Context Protocol > Building MCP with LLMs
• Model Context Protocol > Debugging
• Model Context Protocol > Inspector