Start With a Small Prompt and Build From the Result
This Short is about starting smaller.
A large prompt can feel like preparation, but it can also contain assumptions about an answer I have not seen yet. A smaller first request gives me something concrete to inspect.
I would begin with the task and the output I want. Then read the result and identify the biggest mismatch. Maybe the tone is wrong. Maybe a useful example is missing. Maybe the format needs to change.
Add that correction and try again.
This is iteration, not a guarantee that a short prompt always wins. Some tasks need detailed constraints from the start. The useful habit is to make each addition earn its place by solving an observed problem.
Try it on a low-stakes draft: begin with one clear request, make one correction, and notice what changed.