Pentagon Model
Refine Your Prompt: from good to professional
A practical course for building clear and powerful AI prompts
You have mastered the five pillars. But in reality, a prompt is rarely perfect on the first try — the real skill is not writing a flawless prompt at once, but reading the output and refining the prompt. This is a short loop you repeat until you reach the result you want.
Core principle
An effective prompt is built by iteration, not instant perfection. Run it, read the output critically, spot the weakest pillar, fix only that, and re-run.The refine loop in 4 steps
- Run your prompt and read the full output.
- Ask: where did it fall short? (tone, missing detail, messy shape, wrong assumption)
- Map the gap to its pillar: tone → role/constraints, detail → data, order → task & outputs.
- Edit only that pillar and re-run — do not rewrite from scratch.
First draft
Write a marketing message for a perfume store.After two refines
Act as a marketing expert. For a premium perfume store targeting women before Eid, write a 3-line, elegant, ready-to-send message.Pro tip
The lab in the next lesson is built for this: change one pillar, re-run, and watch the output shift. That is the refine loop in practice.
🤖
Ask the Coach
An AI assistant that answers based on this lesson. Ask freely.
🤖 Coach answers are AI-generated; verify important details.
Check Your Understanding (2 questions)
Question 1
What is the essence of prompt refinement?
💡 Why: Refinement is a loop: read output, diagnose the weakest pillar, fix only it, re-run — not a full rewrite or padding.
Question 2
The reply came formal but you want it friendly — which pillar do you fix first?
💡 Why: Tone is set via Role or Constraints; fix only the weakest pillar, not everything.
The recommended next step unlocks only after the correct answer, and your progress is saved on this device.