ChatGPT MCP Server

A Model Context Protocol (MCP) server that provides Docker management capabilities through a custom GPT interface.

Features

Docker container management through natural language
Built on the Model Context Protocol (MCP)
TypeScript implementation
Containerized deployment
Robust error handling and graceful shutdown
Resource management and port handling
Rate limiting and API key authentication

Setup

Installing via Smithery

To install ChatGPT MCP Server for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @Toowiredd/chatgpt-mcp-server --client claude

Manual Installation

Clone the repository

git clone https://github.com/toowiredd/chatgpt-mcp-server.git
cd chatgpt-mcp-server

Install dependencies

npm install

Create environment file

cp env.example .env
# Edit .env with your configuration

Build the project

npm run build

Running with Docker

Build the container

npm run docker:build

Run the container

npm run docker:run

Or manually:

docker run -d \
  -p 3001:3001 \
  --env-file .env \
  -v /var/run/docker.sock:/var/run/docker.sock \
  chatgpt-mcp-server

Development

npm run build - Build the TypeScript code
npm run watch - Watch for changes and rebuild
npm run inspector - Run the MCP inspector tool

Environment Variables

API_KEY - API authentication key
HTTP_PORT - Server port (default: 3001)
RATE_LIMIT_REQUESTS - Maximum requests per window
RATE_LIMIT_WINDOW - Window size in milliseconds

Resource Management

The server implements robust resource management:

Graceful shutdown on process signals (SIGINT, SIGTERM, SIGQUIT)
Connection tracking and management
Request timeout handling
Port conflict detection
Keep-alive connection management
Active request tracking and graceful completion

Shutdown Process

The server initiates graceful shutdown on process signals
New connections are rejected
Active requests are allowed to complete (with timeout)
Keep-alive connections are closed
Server ports are properly released
Resources are properly released

Error Handling

Port conflicts are detected and reported
Unhandled rejections and exceptions are caught
Network errors are properly handled
Resource leaks are prevented through proper handling
Timeouts ensure the server doesn't hang during shutdown

License

MIT