A good operations prompt isn't a magic trick. It's a checklist you don't have to remember, run by something that never gets bored.
Anchor to one reader
Ask for the reasoning, then the result. When the model shows its steps you can see exactly where to push back.
- Name the audience and the single goal before you generate
- Compare two outputs instead of trusting one
- Have the model flag what it could not cover
Name the failure modes
Lock the parts that never change and leave one slot for the variable. Now every run shares a family resemblance instead of starting from scratch.
Make the output reviewable
Lock the parts that never change and leave one slot for the variable. Now every run shares a family resemblance instead of starting from scratch.
Ask for the reasoning, then the result. When the model shows its steps you can see exactly where to push back.