Publishing tool
Know what to disclose. Fix what reads mechanical.
Answer a few questions about how your book was made and paste an excerpt. Get an instant, free disclosure verdict grounded in Amazon's own KDP guidelines, plus a prose-pattern score — then a $9 report with every flag and fix.
At least 50 characters. Nothing is stored — this is scanned in memory only.
One-time payment, no account.
Free preview
No disclosure required
Based on your answers, no AI was used in the final content, so no AI disclosure applies.
Paste an excerpt to see an instant prose-pattern score.
Frequently asked
Is this an "AI detector"?
No. AI detectors that guess whether text was machine-written are unreliable and easy to fool. This checks for concrete, fixable prose patterns — clichéd filler phrases, uniform sentence rhythm, repeated paragraph openers — that read as mechanical regardless of how the text was produced, and tells you exactly how to fix them.
Do I actually have to disclose AI use to Amazon?
Per KDP's own published guidelines: if an AI tool created the actual text, images, or translation, it's "AI-generated" and must be disclosed — even if you applied substantial edits afterwards. If you wrote the actual content yourself and only used AI to brainstorm, edit, or check grammar, that's "AI-assisted" and needs no disclosure. The free preview below classifies your specific case.
What exactly does the prose scan check?
Four independent signals: density of AI-associated cliché phrases ("delve into," "testament to," etc.), sentence-length uniformity, repeated paragraph openers, and overuse of certain transition words. Each is scored and explained so you know precisely what to revise.
Do you store my manuscript?
No. The excerpt is scanned in memory for the request and never written to a database. The paid report carries your excerpt through Stripe's checkout redirect only, not server storage.
What's in the $9 report that the free preview doesn't show?
The free preview shows the disclosure verdict and your single most severe prose flag. The full report unlocks every flag detected, with the specific detail and fix for each.