Describe what you want in plain language — get back a structured prompt built for the model you're actually using. Already have a prompt? Switch to the Analyzer to see how it scores.
It takes a rough, plain-language request and rewrites it into a structured prompt following the real conventions of whichever AI model you're targeting — chat models get role, context, task, and format structure; image models get the terser, visually-descriptive style they actually respond to.
It's a deterministic checklist of established prompt-engineering practices — role definition, a specific task, context, output format, constraints, examples, and length — not an AI-generated number. The same prompt always scores the same way, and every point is explained.
Yes — it's stored only in your browser's local storage on this device, never sent to or stored on a server.
Yes — every template is written to be copied, have its bracketed placeholder filled in, and used immediately.