I think I agree with all you said. But I can’t let go of a clean codebase yet. It’s stronger than me. And I’m not sure I should. Or at least not yet, not until the agent can better understand what I actually want.

And you’re right, with a bit of direction and further prompting it’s quite easy to get your code in a reasonable shape. So that is what I do now: the agent does all the work but I still whip the output in shape.

Mostly my hope is that at some point soon the agents will start writing clean code. But then again, maybe clean code for a galaxy brained agent is not the most optimal code from a all-in token efficiency perspective.