Drop an Office file below to inspect supported metadata and recoverable content. DOCX inspection includes revision identifiers, stored names and timestamps, tracked deletions and image content outside the displayed crop. Metadata can be incomplete or edited; it does not prove who wrote the text, how it was written or how long it took. Choose supported cleanup options, then review the downloaded result before sharing.
or click to choose · nothing leaves your machine
Revision identifiers. See text grouped by stored ids, not verified human work sessions. Unlisted ids. Body ids absent from the stored list may reflect copying or other document processing; they do not identify a source document. Editing-time metadata. Inspect the reported fields without treating them as measured typing speed. Recoverable content. Unaccepted tracked deletions, hidden/white/1-point text, comments, AutoText, embedded workbooks. Image excavation. Word keeps the whole picture and displays a window onto it, so the cropped-away part is still in the file — along with EXIF, GPS and camera model. Lineage. Drop a second version to see shared sessions and, via stable paragraph ids, exactly which paragraphs were silently rewritten.
Nineteen fixes, from rsid stamps and paragraph ids to tracked changes, comments, invisible runs, EXIF/GPS, sensitivity labels, AutoText, page thumbnails, UNC paths and language tags — plus baking image crops so the off-camera pixels stop travelling. The rewritten file is re-analyzed in front of you, with remaining findings shown. Earlier container/text checks covered 196 supported files; those checks do not establish unchanged rendering. Independent LibreOffice comparison is in progress.