Transparency

A privacy claim is useful only when its scope and limits can be checked.

What is actually reviewed

The project combines three layers: a static external check, a real-browser deep scan and server-side evidence for facts that cannot be proven from outside.

What the deep scan adds

A source-code scan can miss behaviour triggered only after JavaScript runs. The deep scan opens several internal pages in Chromium and records actual runtime requests, third-party hosts, cookies, LocalStorage, SessionStorage, IndexedDB, service workers, external iframes, Fetch/XHR, sendBeacon, WebSockets and selected fingerprinting-relevant APIs.

The scan is deliberately passive: it does not submit forms, create accounts, purchase anything or follow obviously state-changing links.

What even the deep scan cannot prove

A browser can observe browser-side behaviour, but it cannot see private server logs, application logs, provider retention, internal databases or backups. Those areas still require operator information and evidence.

Public verification

Valid badges point to a public record with the domain, verification ID, status and review information.

Shared records

The German and English editions use the same underlying verification records. A language change does not create a second certification.

Published methodology and versioning

The complete weighting, factors, score caps and score levels are published on the methodology page. The current methodology is v1.1.3, while historical records retain the version used at the time.

History and automated re-checks

Published sites keep a review history. External re-checks run periodically. A lower score or a new external finding creates an admin review alert instead of silently replacing the previous state.

Limitations

A review is a point-in-time assessment. Websites and infrastructure can change afterwards; the badge therefore represents the documented status, not an unlimited guarantee.