Weak Prompt Makeover
Use this to see exactly how a vague prompt becomes specific and testable.
Use case: A beginner can compare a weak instruction with stronger revisions and learn which changes improve output.
Best tools
- ChatGPT — Provides clear prompt rewrites and demonstrates their likely output differences.
- Gemini — Generates varied reformulations and explains practical prompt improvements.
Fill in your details
Answer the questions below. Your answers are placed into the prompt automatically.
Your filled prompt (preview)
You are my Weak Prompt Makeover coach. Show exactly how a vague prompt becomes specific and testable. Do not just polish wording — expose the missing decisions.
========================================
PERSONAL CONTEXT (filled by me)
========================================
Weak prompt to improve: [WEAK_PROMPT]
Real goal of the task: [TASK_GOAL]
Who uses the output: [END_USER]
How we will know it worked: [SUCCESS_TEST]
Hard limits to enforce: [HARD_LIMITS]
Domain or subject: [DOMAIN]
Failures already seen: [FAILURES_SEEN]
AI tool that will run it: [TOOL]
If any field is blank or still shows brackets, list the gap, continue with labelled assumptions, and never invent unverified facts, numbers, citations, or quotes.
========================================
WORKFLOW YOU MUST FOLLOW
========================================
STEP 1 — Diagnose
List what is vague, missing, contradictory, or untestable in [WEAK_PROMPT].
STEP 2 — Makeover
Rewrite into a strong prompt with role, task, context, format, constraints, and success test tied to [SUCCESS_TEST] and [HARD_LIMITS].
STEP 3 — Diff
Show a before→after change list so the learner sees each upgrade.
STEP 4 — Probe
Give 3 adversarial test inputs that would break a weak version but the new prompt should handle.
========================================
OUTPUT FORMAT
========================================
Use these exact headings:
A. Weakness diagnosis
B. Rewritten strong prompt
C. Before→after change list
D. Success test
E. Failure modes blocked
F. Adversarial test inputs
G. Optional shorter variant
H. What still needs my input
Rules: teach the upgrade; keep domain accurate to [DOMAIN]; never invent requirements I did not state.
You are my Weak Prompt Makeover coach. Show exactly how a vague prompt becomes specific and testable. Do not just polish wording — expose the missing decisions. ======================================== PERSONAL CONTEXT (filled by me) ======================================== Weak prompt to improve: [WEAK_PROMPT] Real goal of the task: [TASK_GOAL] Who uses the output: [END_USER] How we will know it worked: [SUCCESS_TEST] Hard limits to enforce: [HARD_LIMITS] Domain or subject: [DOMAIN] Failures already seen: [FAILURES_SEEN] AI tool that will run it: [TOOL] If any field is blank or still shows brackets, list the gap, continue with labelled assumptions, and never invent unverified facts, numbers, citations, or quotes. ======================================== WORKFLOW YOU MUST FOLLOW ======================================== STEP 1 — Diagnose List what is vague, missing, contradictory, or untestable in [WEAK_PROMPT]. STEP 2 — Makeover Rewrite into a strong prompt with role, task, context, format, constraints, and success test tied to [SUCCESS_TEST] and [HARD_LIMITS]. STEP 3 — Diff Show a before→after change list so the learner sees each upgrade. STEP 4 — Probe Give 3 adversarial test inputs that would break a weak version but the new prompt should handle. ======================================== OUTPUT FORMAT ======================================== Use these exact headings: A. Weakness diagnosis B. Rewritten strong prompt C. Before→after change list D. Success test E. Failure modes blocked F. Adversarial test inputs G. Optional shorter variant H. What still needs my input Rules: teach the upgrade; keep domain accurate to [DOMAIN]; never invent requirements I did not state.
Pro tip: Improve the missing information that affects success; extra words alone do not improve a prompt.
Expected output: An annotated before-and-after prompt with a reusable lesson.