Bharadwaj-Valluri-cmm/mcp-server-demo

PDLC MCP Server Demo

This project demonstrates a simple MCP (Model Context Protocol) server implementation in Python. It provides tools, resources, and prompts for interacting with the server via stdio.

Features

• Palindrome Checker: Checks if a string is a palindrome.
• Addition Tool: Adds two numbers (with an extra +3 for demonstration).
• GitHub Connector: Fetches GitHub user details.
• Greeting Resource: Returns a greeting for a given name.
• Prompt Generator: Generates greeting prompts in different styles.
• Math Expression Evaluator: Evaluates mathematical expressions.
• Code Generator: Generates Python code for a given task.

File Structure

PDLC/
├── main.py                # Main MCP server implementation
├── __pycache__/           # Python cache files
└── mcp-server-demo/
    ├── pyproject.toml     # Project configuration
    ├── README.md          # Demo README
    └── __pycache__/

Getting Started

1. Install dependencies
   • Ensure you have Python 3.8+ installed.
   • Install required packages:

     pip install requests
     

2. Run the MCP server
   • From the project root:

     python main.py
     

   • Or use the configured command in .vscode/mcp.json for stdio mode. For example, if you have uv.exe installed and configured:

     C:\Users\ei1eipe\.local\bin\uv.exe run main.py
     

   • You can also run other configured servers, such as:

     C:\Users\ei1eipe\.local\bin\uv.exe
     

Customization

• Add new tools by decorating functions with @mcp.tool().
• Add resources with @mcp.resource().
• Add prompts with @mcp.prompt().

Connecting MCP with Copilot

You can integrate your MCP server with GitHub Copilot or similar AI coding assistants to automate code generation and enhance developer productivity. Here’s how you can connect and use Copilot with your MCP server:

Integration Steps

1. Ensure Copilot is enabled in your IDE (e.g., VS Code).
2. Start your MCP server using one of the methods described above.
3. Use Copilot to interact with MCP tools by writing prompts or code that calls the MCP server’s exposed functions (such as add, connect_github, etc.).
4. Leverage Copilot for code generation by using the generate_code tool, which is designed to work with Copilot-like models:

result = generate_code("Sort a list of numbers", lang="python")
print(result)