Fair questions.
Straight answers.

You give it a privacy policy — a URL or the raw text — and it reads the whole thing in seconds. It isolates the clauses that actually hurt you, translates each into plain English, ranks them worst-first, and pins every single one to the verbatim sentence in the policy that proves it. No summary fluff, no wall of text.

Nine known traps: selling or sharing your data, cross-site tracking, no way to delete your data, perpetual licenses over your content, forced arbitration, silent policy changes, biometric or sensitive data collection, children's data handling, and deliberately vague euphemisms. Each is scored critical, high, or medium.

No. The scan is stateless — the policy is analyzed in the moment and discarded. We don't keep the text, the URL, or a profile of you. A privacy tool that harvested your data would be the punchline of a very bad joke.

It can miss things, but it is built not to invent them. A flag is only shown if its supporting quote is verified word-for-word against the source document. No matching quote, no flag. That's the house rule — accusations require receipts.

It means none of the nine traps matched — which is a genuinely good sign, not a guarantee. Policies can hide problems in ways no scanner catches, and they change. Re-scan when a service emails you that dreaded “we've updated our privacy policy.”

No. It's an early-warning system, not a lawyer. It tells you where to look and what the words mean in plain English. For decisions with real stakes, take the flagged clauses to a professional.