An open request for use cases, snippets, and expected behavior. Feel free to post here.

Not sure there's enough that can be done with code to train a mode.