claude-mcp

Claude MCP (Monitoring Control Panel)

A simple, lightweight server that allows Claude to analyze and improve Python code quality in The Everything App project.

Features

• 🚀 Easy to install and use with NPX
• 🔍 Python code quality analysis with Pyright
• 🛠️ File operations (read, write, search)
• 🌐 GitHub integration
• 🧠 Works with Claude Desktop App

Installation

Install the package globally:

npm install -g claude-mcp

Or run directly with NPX:

npx claude-mcp start

Quick Start

1. Start the server:

   npx claude-mcp start
   

2. The server will run on port 3333 by default.

3. You can now interact with the server using Claude Desktop App or any HTTP client.

Usage with Claude

When using Claude Desktop App, ask Claude to analyze your Python code:

Claude, please analyze the Python code in my project using the local MCP server at http://localhost:3333

Claude can then use the server's API to:

• Check for Python code issues
• Suggest fixes
• Create PRs with improvements

API Endpoints

Python Code Quality

• POST /api/python/analyze - Analyze a Python file or directory
• POST /api/python/fix - Apply fixes to a Python file

File Operations

• POST /api/files/list - List files in a directory
• POST /api/files/read - Read a file
• POST /api/files/write - Write to a file
• POST /api/files/search - Search for files

GitHub Integration

• GET /api/github/repos - List GitHub repositories
• POST /api/github/content - Get repository content
• POST /api/github/update - Create or update a file
• POST /api/github/pr - Create a pull request

CLI Commands

• claude-mcp start - Start the server
• claude-mcp analyze <path> - Analyze a Python file or directory
• claude-mcp init - Initialize a new Claude MCP project

Environment Variables

Create a .env file in your project root with the following variables:

PORT=3333
GITHUB_TOKEN=your_github_token

Integration with The Everything App

This tool is specifically designed to help Claude analyze and improve Python code quality in The Everything App project, focusing on:

• AWS Lambda functions
• SageMaker models
• Data transformation scripts
• API integrations

License

MIT