headerless

Strips format-defining header fields from binary assets at build time, and puts them back in memory at runtime.

Not encryption, not DRM — it raises the cost of extracting assets from a shipped app from unzip to reverse-engineer. Anything visible to the renderer stays visible to a determined attacker; that is by design.

⚠️ WIP — API is not stable and this package is not published yet.

How it works

The cut fragment must be the key that reconstitutes the file. That single rule drives everything else:

• A fixed-length cut is rejected. PNG signatures and JPEG SOI/APP0 are identical across files, so a restorer just writes the constant back.
• Cuts are per format and target the fields that vary per file: PNG IHDR data and CRC, the WebP frame header, JPEG SOF0.
• Constant runs are removed from the file but not stored in the fragment — they are regenerated on restore.
• Formats without a parser fail loudly instead of falling back to a fixed cut.

Because the fragment is a per-file key rather than a shared constant, publishing the algorithm does not weaken it (Kerckhoffs). Only the fragments are secret.

Status

| | | | --- | --- | | detect() | done | | locate() / strip() / restore() | not implemented |

The design rationale is documented in the repository README.

License

MIT