I think that’s because across experience levels there really isn’t much to teach regarding “how to use AI”.

The only grind is still just do you know the subject matter, tools and frameworks, etc. because that’s the knowledge that lets you look at the output and say “this is fine” or “this is absolutely the worst implementation I’ve ever seen try again but do X”.

In my experience adding on multiple layers of controls etc aside from basic skills is essentially as useful as a ring of salt and a live offering.

My workflow these days is basically ultracode + PRs. I do control the features that go but barely need to check the code due to the amount of checks: lint, TS, Local AI visual Model for visual check for screen frame comparison (Mac M5 48GB mem), Pixel/Screenshot regression test for basic UI, code coverage/c8 to avoid trash code without tests, etc. etc.

Essentially the tests are just for the AI to remember what they did (as in the description + break if they change smth unintentionally).