Prompt analyzer
Analyze your prompt, catch problems, and improve it for each AI.
Analyze, fix and improve a prompt.
Prompt of the dayTests-first bug huntCoding
Prompt of the day
I have a bug I can't reproduce consistently: """[bug description + environment]""" Before proposing a fix: 1. List 5 root-cause hypotheses, ordered by likelihood. 2. For each one, write a test (with code) that confirms or rules it out. 3. Only then, propose the minimal fix assuming the most likely hypothesis. 4. End with how to prevent a bug like this in the future.
I'm writing for…
What is your prompt for?
Prefer talking? Dictate your prompt with the mic and we'll turn it into text you can review.
Choose a prompt type to enable attachments.
Optimized prompt format
What the prompt analyzer checks
Every prompt gets a 0–100 score on six criteria, weighted by the purpose you pick. Then you see what to fix first and an optimized version, scored again with the same rules.
- Clarity. Is there a stated goal and enough words to say what is wanted? Very short or goal-less requests score low.
- Context. Does the prompt include the material to work on (inputs), say who the answer is for (audience), or give examples?
- Constraints. Are there explicit limits such as tone, length, timeframe, region or rules the answer must respect?
- Output format. Does it say what shape the answer must have (list, table, JSON, steps), and what a good answer looks like?
- Verifiability. Can the answer be checked? Success criteria, error details and reproduction steps count here; it weighs most for debugging and code.
- Safety. Starts at 100 and drops for text that looks like a prompt injection, for contradictory instructions and for a language mismatch with the request.
See before → after examples with their scores · How Promptea works
FAQ
What makes a prompt score high?
A clear goal, the context the model cannot guess, explicit constraints, the exact output format and a way to check the answer. A short prompt can score well too, as long as it says what is needed; padding does not add points.
Can I check prompts for code, JSON or images?
Yes. Pick the purpose — text, study, code, data/JSON, image, marketing, translation or summary — and the analyzer applies the checks that matter for it, like reproduction steps for a bug or a schema for JSON.
Why does the same prompt always get the same score?
The score comes from fixed rules, not from a model's opinion, so you can change one line, analyze again and see exactly what that line was worth.