Feature request: LLM / GPT integration

[evelant]

It would be awesome to have an option for thefuck to query an LLM when there's not a clear match. A lot of models are pretty good at correcting terminal commands and almost nobody remembers all the arguments and syntax to everything. Bonus if you could query it upfront, maybe with a fun alias like howthefuck do I kill all gradle processes or howthefuck find all files in this directory containing "shit"