Repair instead of regenerate
Diagnose a weak output and apply the matching repair move.
The first answer is a diagnostic, not a verdict. Skilled users expect two or three rounds and treat each disappointing output as information about what the prompt failed to specify. The unskilled pattern is to regenerate repeatedly hoping for luck; the skilled pattern is to name the defect and add the missing constraint. Regeneration without a change is the definition of wasted effort. There are four repair moves, and matching the move to the defect is the whole skill. If the output is too general, add specifics and a narrower task. If it is off-voice, add an example. If it is incomplete, ask for the missing element by name and keep what you liked. If it is wrong, correct the fact explicitly and tell it to use your version. The instruction "keep paragraphs one and three, rewrite two to include the pickup deadline, shorter" is worth more than any full restart. End every productive round by harvesting. Once the third attempt lands, fold what you learned back into a saved prompt so round one is that good next time. Iteration should be a one-time cost per task type, not a permanent tax. When you notice yourself making the same correction twice, that correction belongs in the prompt, not in your hands. Watch for the moment you feel like starting over. That impulse almost always arrives when the output is mostly right, which is exactly when a targeted repair is cheapest and a restart is most wasteful. Keeping the good parts explicitly is an instruction, not an assumption. Say which paragraphs to preserve, or the next version will quietly rewrite them. A grant writer tracked her rounds for a month and found that 80 percent of her tasks reached usable quality in two rounds once she began naming defects instead of regenerating. Her saved prompts absorbed the rest.
Key takeaways
- There are four repair moves, and matching the move to the defect is the whole skill.
- End every productive round by harvesting.
- Targeted revision preserves what already works.
Sign in to complete activities, save progress, and use the AI Coach.
Evergreen instructional material; no time-sensitive claims to source.
Evergreen curriculum written from KleinHub teaching material — no time-sensitive claims.
AI-generated lesson, reviewed by Education Review and approved through the KleinHub approval queue before publication.
Sign in to complete this lesson, track progress, and use the AI Coach.
Sign in to start lessonCreate account