alorse/copilot-to-api ? reverse-engineered prompt

Build me a small tool that lets me use GitHub Copilot like an OpenAI style API.

I want a simple setup where I can add my Copilot access token once, then run either a local web server or a command line client and send chat requests to it just like I would with OpenAI. It should support listing models, chat completions, and streaming responses, and work with both Node and Python so I can pick whichever is easier.

Please include a clear way to configure the token, a basic health check, and a few example requests I can try right away from curl and from an OpenAI client library. If you need to check anything current in the docs or endpoints, look it up online first.

Keep it easy to run locally, with simple start commands and helpful error messages if the token is missing or expired.