This page downloads a media file and reads its bytes in your browser — no data is transmitted to any server, and the one that published the file plays no part in the result. It lists which metadata containers the file actually holds, so a Scrub declaration such as “metadata stripped” can be checked rather than believed. The code doing the reading is this file, about two hundred lines, no dependencies.

Limits: this inspector can only check transformations that leave a trace in the file (metadata removal). It cannot tell whether faces were blurred or a voice altered — there is no original to compare against. A file that is clean today may be replaced tomorrow at the same URL; the declared hash, when present, is what ties the declaration to these exact bytes. You are trusting this page's code rather than the publisher's word: less than a promise, more than nothing. The exiftool utility can also tell you about the metadata in question.