Here's exactly what happens in the seconds between pasting a link and reading your verdict.
We collect 200+ facts about the site from trusted public sources. Age, security, owner, reputation and more.
Each fact becomes a signal: good, neutral or bad. Points are calculated depending on how serious they are.
AI weighs it all in context and writes the verdict. A perfect site is near 100; serious problems sink the score fast.
Every site is scored 0–100 in six areas. Areas that protect your money and identity count the most.
Is the connection encrypted, and is the site clean on malware and phishing scanners?
Who owns the site, how long has it existed, and are they hiding?
What do blocklists, web history and reviews say about it?
Can you actually find out who runs it and how to reach them?
Are the legal basics in place, are they valid, is the company real?
Is the site built on solid ground — proper hosting, DNS and email setup?
Each fact we collect is classified into one of four kinds. Only warnings and bad signals cost points.
Things working in the site's favour. They don't add points; trust is the default.
Useful context that doesn't change the score either way.
Worth knowing about, but common on otherwise honest sites.
Red flags strongly linked to scams and fraud. These cost real points.
Every site starts at 100. Problems subtract points based on severity. One severe finding is enough to sink a score on its own.
Tidiness issues with no real risk: a missing favicon, no social links.
−1 to −3Sloppy but survivable: weak security headers, no cookie notice.
−4 to −9Genuine trust problems: hidden ownership on a store, very young domain, no contact details.
−10 to −24Strong scam indicators: blocklist hits, malware detections, brand impersonation.
−25 to −60The same fact can be harmless on one site and alarming on another. Our model reads each signal in the context of what the site actually is.
Not great, but a hobby blog isn't asking for your card number. Treated as a cosmetic issue, a few points at most.
A store taking payments with no privacy policy is a serious problem. The same fact is now a major deduction.
The AI never invents facts. It only weighs the signals our checks actually found, and every report shows you those signals so you can see the working for yourself.
We don't guess. Every signal is collected live from authoritative public sources at the moment you check. We don't publish the full list, to make the system harder to game.
Anyone can re-scan a site at any time. Fix the problems, and the score recovers automatically.
Nobody can pay for a better score. Badges show the live score, whatever it is.
We can measure signals, but not intentions. A high score means low risk. It's not a promise, and it isn't legal or financial advice.
The best way to understand the score is to read a real report.