ChatGPT-MCP Server logo

ChatGPT-MCP Server

by automateyournetwork

This is a Model Context Protocol (MCP) stdio server that forwards prompts to OpenAI’s ChatGPT (GPT-4o). It is designed to run inside LangGraph-based assistants and enables advanced summarization, analysis, and reasoning by accessing an external LLM.

View on GitHub

Last updated: N/A

What is ChatGPT-MCP Server?

The ChatGPT-MCP server is a Model Context Protocol (MCP) server that acts as an intermediary between a LangGraph-based assistant and OpenAI's ChatGPT (GPT-4o). It allows the assistant to leverage the advanced capabilities of ChatGPT for tasks like summarization, analysis, and reasoning.

How to use ChatGPT-MCP Server?

The server can be deployed using Docker or run manually. It exposes a single tool, 'ask_chatgpt', which accepts text as input ('content') and sends it to ChatGPT for processing. It can be integrated into a LangGraph pipeline by configuring the server command and arguments.

Key features of ChatGPT-MCP Server

  • Exposes ChatGPT as a tool for LangGraph agents

  • Supports advanced summarization and analysis

  • Enables natural language reasoning

  • Uses a simple stdio interface

  • Dockerized for easy deployment

Use cases of ChatGPT-MCP Server

  • Summarizing long documents

  • Analyzing configuration files

  • Comparing options

  • Performing advanced natural language reasoning

FAQ from ChatGPT-MCP Server

What is the purpose of this server?

It allows LangGraph agents to access the advanced capabilities of ChatGPT for tasks like summarization and reasoning.

How do I deploy this server?

You can deploy it using Docker or run it manually using Python.

What is the 'ask_chatgpt' tool?

It's the tool exposed by the server that sends text to ChatGPT for processing.

What dependencies are required?

The server requires the openai, requests, and python-dotenv libraries, which are installed during the Docker build.

How do I configure the OpenAI API key?

You can set the OPENAI_API_KEY environment variable in a .env file or export it manually.