What Made My Prompts Succeed or Fail
“You can’t calm the storm, so stop trying. What you can do is calm yourself. The storm will pass.” — Timber Hawkeye
Typhoon Uwan arrived quietly at first, slow winds, then steady rain, until it demanded that everything pause.
Plans changed. Runs postponed. Schedules rearranged.
It felt strangely familiar.
Because that’s exactly what happens in prompt engineering. You think you’ve built something solid, a clean system, a clear structure, until it’s tested by unpredictability.
This week, I learned that the prompts that worked best were the ones built like storm-ready houses: clear purpose, firm roles, strong framing. They held up even when the wind shifted.
The ones that failed? They were vague, scattered, unanchored, and they collapsed at the first sign of confusion.
Funny how both storms and AI remind us of the same truth: you can’t control the chaos, but you can design for it.
The tabular format I used for my running programs and planning prompts became my structure, my roof and beams. It kept things steady when the system felt overwhelming.
And as Uwan moved away, leaving the city quiet again, I realized this study isn’t just about AI. It’s about learning how to think, plan, and adapt, whether it’s a storm outside or one in your own process.
OGRap 