IntactFile — sample corrupted files =================================== These files are INTENTIONALLY BROKEN. They exist so you can test IntactFile in about 30 seconds without risking a file you care about — and, more importantly, so you can prove the "0 bytes uploaded" claim for yourself: open your browser's DevTools Network tab, drop one of these into the widget on intactfile.com, and watch. You will see the page's own assets load, and then nothing carrying your file — no matter which sample you pick. They are tiny, synthetic containers (a few hundred bytes to a few KB) generated by our own corpus tools. There is no real footage, no real document text and no personal data inside — just the minimum valid structure with a small dummy payload, then a specific, deliberate kind of damage. What each file is, and what IntactFile does with it: sample-truncated.mp4 (~1.3 KB) An MP4 that was cut off before it finished writing — the tail of the media data never made it to disk (a "truncated" recording). The index at the front survives. IntactFile diagnoses this as "truncated-mdat", rebuilds the container around the frames that DID survive, and hands back a playable file. It is honest about the cut: the frames after the truncation point are physically gone and are not invented. NOTE: a file whose index was never written at all ("missing moov") is a different case — it needs a healthy same-camera reference clip to reconstruct, so it is not a single-file drop-and-repair demo. This sample is the truncation case, which repairs from the one file alone. sample-broken.zip (~5.1 KB) A ZIP archive cut short before its end — the classic "unexpected end of archive" error. The end-of-archive record and the last entry are gone. IntactFile scans the local entries that survived, verifies them, and rebuilds a fresh, valid archive from the ones that are intact. With this sample that is 3 of the files; the entry caught mid-cut is reported, honestly, as damaged beyond repair rather than silently dropped. sample-damaged.pdf (~0.6 KB) A PDF whose cross-reference table (the index a viewer uses to find each object) has been destroyed — this is the "the file is damaged and could not be repaired" case many PDF viewers report. The page objects themselves survive. IntactFile rebuilds the xref and returns an openable PDF that re-scans as structurally clean. Everything happens locally, in your browser tab. Nothing here is uploaded.