The 4 parts of a good prompt

The 4 parts of a good prompt

Why it matters: A structured prompt reduces ambiguity and makes outputs repeatable.

Goals

• Name the 4 parts of a good prompt.
• Rewrite a weak prompt into a strong one.
• Add criteria so output is testable.

Key definitions

• Context: Relevant background so the model knows the situation and constraints.
• Task: What you want produced (draft, summary, table, plan) in clear terms.
• Constraints: Rules: format, length, tone, excluded items, do/don’t, must-include.
• Success criteria: How you judge quality (checklist, rubric, examples of good/bad).

Workflow (step-by-step)

1. Write one sentence of context (who, what, why).
2. Write the task as an output request (deliverable + format).
3. Add 3–7 constraints (tone, length, must include/exclude, structure).
4. Add success criteria (checklist + 1 good / 1 bad example).
5. Ask for clarification questions when missing inputs.

Example (good vs bad)

✅ Good: Prompt includes context + task + constraints + checklist; the output is structured and easy to verify.

❌ Bad: Prompt says “Write something about X” with no audience, constraints, or success definition.

Checklist

• Context is 1–3 sentences.
• Task specifies output format.
• Constraints include must-include and must-not.
• Success criteria is explicit.

Metrics / criteria

• Measurable: You can say whether the output is correct/complete (pass/fail or a score).
• Effort: You can produce the result in 15–30 minutes using the workflow.
• Clarity: Another person can run your prompt and get a similar outcome.
• Repeatability: 2 runs produce outputs that meet the same checklist.

Common mistakes + fixes

• Pitfall: Constraints missing or conflicting. Fix: Keep constraints short; remove contradictions; add a priority order.
• Pitfall: No examples. Fix: Add a small “good vs bad” snippet to anchor style and structure.
• Pitfall: Vague instructions. Fix: Add constraints, examples, and a success checklist.
• Pitfall: One-shot prompting. Fix: Iterate: draft → critique → revise → verify.
• Pitfall: No validation. Fix: Add QA steps, tests, and explicit “what counts as done”.

Student tasks

1. Apply the workflow to one real work task related to the 4 parts of a good prompt.
2. Create a small artifact: prompt, checklist, rubric, table, draft, or decision note.
3. Run one quality check and record what changed after the check.
4. Write a 3-sentence reflection: what worked, what failed, and what you will reuse.

Useful external sources

• OpenAI prompt engineering guide: https://platform.openai.com/docs/guides/prompt-engineering — Prompt structure, iteration, constraints, and evaluation habits.
• NIST AI Risk Management Framework: https://www.nist.gov/itl/ai-risk-management-framework — Risk framing, governance language, and verification mindset.
• OWASP Top 10 for Large Language Model Applications: https://owasp.org/www-project-top-10-for-large-language-model-applications/ — Practical AI safety risks such as prompt injection, leakage, and insecure output handling.
• Google People + AI Guidebook: https://pair.withgoogle.com/guidebook/ — Human-centered AI design, user control, feedback, and failure modes.

Bibliography

• OpenAI prompt engineering guide. https://platform.openai.com/docs/guides/prompt-engineering
• NIST AI Risk Management Framework. https://www.nist.gov/itl/ai-risk-management-framework
• OWASP Top 10 for Large Language Model Applications. https://owasp.org/www-project-top-10-for-large-language-model-applications/
• Google People + AI Guidebook. https://pair.withgoogle.com/guidebook/