Free demo · Prompt Engineering · Lesson 2 · The advanced prompt framework

Seven building blocks of a great prompt

A structured way to craft prompts that gives more accurate, relevant and creative outputs. Watch one complete prompt grow, block by block:

  1. 1Purposeend goal · kind of help
  2. 2Contextaudience · tone · domain
  3. 3Output formatbullets · steps · table
  4. 4Constraintslimits · exclusions
  5. 5Interactions“ask me questions”
  6. 6Examplessample output
  7. 7Iterationexpand · simplify · fix
You are a climatologist and science teacher. Write an educational blog on the effects of global warming on human life for high-school students, in a warm but factual tone. Use a title, three short sections with headings, and a 5-point summary at the end. Keep it under 250 words, avoid ocean ecology, and explain any technical term you use. Summary points should look like: “Heatwaves are longer, so schools now start earlier in summer.” Ask me questions if anything is unclear before you start. After the draft, wait for my feedback.
  1. 1
  2. 2
  3. 3
  4. 4
  5. Enrol

Tip: press → or Space to step through, ← to go back, R to replay.

About this lesson

The advanced prompt framework breaks a prompt into seven building blocks: define the purpose, set the context (audience, tone, domain), choose the output format, add instructions and constraints, enforce interactions so the AI asks questions first, give examples of the output you want, and iterate on the answer. You rarely need all seven: a quick question needs purpose and context, while a report or a lesson plan benefits from every block.

Key points

  • Purpose, context and output format
  • Constraints, interactions and examples
  • Iteration: expand, simplify, fix