Five-line brief & improvement loop
From Lesson 11: a prompt is context that reshapes the answer — briefing quality drives output quality almost linearly. Fill in the five lines below before you ask, and use the loop on the back for anything worth a second pass.
The five-line brief
1. Objective — what you want, and for whom:
2. Context — what it cannot know without you: your situation, data, audience:
3. Constraints — length, tone, format, what to avoid:
4. Example — a sample of the shape or voice you want (patterns beat descriptions):
5. Uncertainty rule — how it should handle not knowing something ("say unknown rather than guess"):
The controlled improvement loop (one brief rarely finishes a job — endless re-prompting is the expensive way to learn that)
- Define the objective (line 1 above).
- Get an initial attempt.
- Evaluate it against explicit criteria — written before you look.
- Name the gaps, specifically.
- Request targeted improvement of those gaps only — never "make it better".
- Verify important claims independently. AI critique of AI output is a drafting aid, never verification.
- Keep what you learned — save this brief and your criteria for reuse.
- Stop when another round would add less than it costs. That's a decision, not a failure.
My stop rule for this task — good enough looks like:
Record (for a brief worth reusing)
Task: · Rounds run: · Saved as template?
education.yougenius.ai · The Prediction Machine · free to copy and adapt within your organisation.