Paste any prompt you'd send to an LLM and get a clarity score, breakdown of what's working, and concrete suggestions to get better results. 100% client-side — nothing leaves your browser.
LLMs respond to what you actually write, not what you mean. A vague prompt like "write something about AI" gives a generic result. A structured prompt with a role, format, and examples gives a usable result.
The analyzer checks 5 categories: Role (does the model know who it's being?), Format (is the output structure defined?), Examples (did you show what you want?), Clarity (any vague or weak words?), and Constraints (tone, length, audience specified?).
Tip: A score of 7+ usually means a usable prompt. Under 5 and you'll get generic results. Try the example buttons to see how different prompts score.