MCP Server Deployment Demo

This project is a demonstration of how to create and deploy a simple server that adheres to the Model-Context-Protocol (MCP). It provides a basic tool that can be called by an MCP client, such as an AI model like Claude.

Overview

The Model-Context-Protocol (MCP) is a specification for communication between a large language model (or other AI agent) and a local development environment. It allows the model to access local context and execute tools securely.

This project implements a simple MCP server using the mcp Python library. The server exposes a single tool, add, which takes two integers and returns their sum.

Features

• A simple FastMCP server implementation.
• A basic add(x: int, y: int) -> int tool.
• Packaged as a Python project using pyproject.toml and setuptools.
• Ready for installation and deployment.

Project Structure

• pyproject.toml: Defines project metadata, dependencies, and build configuration. The [project.scripts] section defines the mcp-server command.
• src/mcpserver/deployment.py: Contains the core logic for the MCP server, including the definition of the add tool.
• src/mcpserver/__main__.py: Provides the main entry point to run the server from the command line.

Getting Started

Prerequisites

• Python 3.12+
• uv (a fast Python package installer and resolver)

Installation

1. Clone the repository:

   git clone https://github.com/AdarshChaudhary03/mcp-server-deployment.git
   cd mcp-server-deployment
   

2. Install dependencies: Create a virtual environment and install the project in editable mode.

   uv venv
   source .venv/bin/activate
   uv pip install -e .
   

Running the Server

Once installed, you can run the MCP server using the script defined in pyproject.toml:

mcp-server

Claude Integration

To use this server as an MCP host for a client like Claude, you need to update your Claude configuration file (claude-config.json). Add the following entry to the mcpHosts section.

This configuration tells the Claude client how to launch and communicate with your MCP server. It uses uvx to run the server directly from the Git repository, ensuring you are always using the latest version.

Note: Make sure to replace the command and directory paths with the correct absolute paths for your system.

"MCPServer": {
  "command": "/Library/Frameworks/Python.framework/Versions/3.13/bin/uvx",
  "args": [
    "--directory",
    "/Users/adarshchaudhary/Desktop/ai-projects/mcp-servers/mcp-server-deployment",
    "--from",
    "git+https://github.com/AdarshChaudhary03/mcp-server-deployment",
    "mcp-server"
  ]
}