How the ZIP repair tool works
Use the tool to inspect one ZIP archive at a time and recover entries whose compressed data can still be read. It first checks the central directory, which is the index at the end of a ZIP. It checks whether each recorded offset points to a local file header with the PK\x03\x04 signature. If the directory is missing or unusable, it scans the raw file for local file headers, the records that introduce individual entries.
The raw scan can infer an entry boundary from the next ZIP signature when a usable size is not available. Such entries can be marked with guessed size. The parser handles Stored (0) and Deflate (8) entries. Encrypted entries are excluded because the tool does not guess passwords.
The result is a newly assembled ZIP, not a byte-for-byte repair of the damaged file. Rebuilt output uses Deflate compression and includes only entries that the tool recovered and added successfully.
Understand recovery statuses and results
Use the findings and Entries table to check where each entry came from and what happened during recovery:
- Recovered · verified means the entry decompressed and its CRC32 checksum matched the recorded value when that value was available for comparison. A recorded CRC32 value of zero is treated as verified without an explicit comparison.
- Recovered · unverified means the entry decompressed, but its recorded CRC32 did not match. It may still be included in the rebuilt ZIP.
- Encrypted · skipped means the entry was detected but not decrypted.
- Unrecoverable - the entry could not be decompressed or uses an unsupported compression method.
Raw-scan candidates may show the visible guessed size badge when the entry boundary was inferred. The JSON recovery report includes the recovery findings and entry details, including CRC-related status information.
Repair a damaged ZIP
- Use Drop a damaged ZIP here or select browse file to choose one .zip archive.
- Review Recovery findings and the statuses in Entries.
- Select Download rebuilt ZIP when a rebuilt archive is available.
- Select Download recovery report to save the JSON findings, whether or not a rebuilt ZIP was produced.
Limits and privacy
- The hard scan limit is 300 MB. A file above that limit produces a recovery report with no rebuilt archive.
- The parser has no ZIP64 parsing logic. ZIP64 archives or ZIP64 metadata are unsupported or unverified.
- A recovered entry may still be unverified and can still be included in the rebuilt ZIP.
- An unnamed entry may appear as recovered in the report, but it is not added to the rebuilt ZIP.
- Recovered entries may fail during output archive assembly. If that happens, no rebuilt ZIP is offered, but the recovery report remains available.
- The tool does not guess passwords or scan the archive for malware. Treat recovered files cautiously before opening them.
- Scanning, recovery, and rebuilding happen locally in the browser. The original file is not modified.
Frequently asked questions
Can this tool recover any corrupted ZIP file?
No. It only recovers entries whose compressed data is still physically present and intact enough to decompress - by reading the central directory where possible and falling back to a raw scan for local file headers where it isn't. If an entry's data is genuinely missing, overwritten, or too badly damaged to decompress, it is listed as unrecoverable with a reason, not silently skipped or faked.
Does it crack or bypass passwords on encrypted entries?
No. Encrypted entries are detected and listed in the report but never decrypted, guessed, or brute-forced. They are excluded from the rebuilt archive rather than being tampered with.
Is a corrupted archive checked for malware?
No. This tool only performs structural recovery. A corrupted or recovered archive is not scanned for malicious content and should not be treated as safe just because it opens or its entries pass CRC verification.
Does it modify or overwrite my original file?
No. The source file is read once into browser memory and never written to. Any recovery is written out as a brand new ZIP you download separately - the original stays exactly as it was.
Does it upload my archive anywhere?
No. Scanning, recovery, and rebuilding all happen locally in your browser. Nothing is sent to a server.