Privacy
Your EPUB is treated as temporary scan input.
EPUB Fixer needs the file to run EPUBCheck and inspect package paths. The report uses paths and error summaries, not manuscript text.
What the scan uses
Enough to show the report, not your book text.
- Uploaded EPUB files are used only to run the scan, identify structure errors, and check whether a safe fix passes EPUBCheck again.
- The upload is processed server-side because EPUBCheck and package inspection need access to the EPUB package.
- Event logs store metadata such as file size bucket, error class, issue count, request path, and request ID. They do not store book text.
- Uploaded files are temporary and are designed to expire after 24 hours.
- EPUB Fixer does not use uploaded manuscripts to train models and does not fetch remote resources from inside the book.
What you can do
Use the result page and request ID when you need help.
- Use the result page delete action if you want the uploaded EPUB removed before automatic expiry.
- Do not upload an EPUB if you are not authorized to test or share it for validation.
- Do not upload the file if you need a local-only or offline check.
- Use the request ID when asking about a scan so support can find metadata without needing manuscript text.
- Do not send full book content by email when a path, error code, request ID, or screenshot is enough.