The opposite factor that is required for a ChatGPT plugin is an OpenAPI or Swagger file - these are tedious to put in writing - and it has been identified to me that if I’d used FastAPI, I wouldn’t have wanted to write down one at all. But one of the really nice things you can do with ChatGPT is paste in your Flask code, and it will generate an OpenAPI file for you. I’m simply working this plugin regionally, I don’t have any plans to publish it as that will require a lot more plumbing. For developing local plugins, it is advisable have a server working on localhost. Fortunately, we are able to use the wonderful VSCode assist for distant improvement. I’m connecting by means of to the Pi from my desktop machine, and when i launch the server, VSCode tunnels by way of to the Raspberry Pi and exposes the server on my local machine. ChatGPT seems at the /.properly-identified/ai-plugin.json and makes use of that to get the title, description, logo, and OpenAPI file. That’s all it has to work with to determine what to do with the API. We are able to get some perception into how it really works below the hood. We can just ask ChatGPT to indicate us its pondering.