For now, it injects an 'ask' command in to your typer CLI, which you can use to perform standard model queries via the OpenAI assistant API. These queries have the full typer CLI exported to them as functions. There are examples provided of how the assistant can chain these calls together for you, and do some very cool things.

Additionally, the underlying assistant objects are exposed. This part needs a lot of work still to get it to where I think it needs to be - there is a roadmap in the README.md. I'm paying particular attention to the lifecycle of assistants and doing things like intelligent incremental updates when the underlying assistant definition changes, but that's all a bit rough right now. Again, details in the README.md.

This is my first PyPI published package and I would very much appreciate any helpful tips on how to make this project a success; I think it can do a lot more than it's currently doing, and I would like to commit to growing it.

(Also: I am looking for full-time remote work in backend, tooling, or devops!)