Privacy beyond the pixels.
Every photo you share carries more than the image — EXIF, XMP, and IPTC metadata can silently embed exact GPS coordinates, altitude, timestamps, and device information. Most people never see it. Attackers actively look for it.
MetaVeins is an intent-aware metadata privacy layer for images (and eventually video/audio/documents). Instead of blindly stripping everything, it classifies metadata by purpose and lets the user choose how much to keep, protect, or remove — based on why they're sharing the file.
Status: early-stage / architecture phase. This repo exists to build MetaVeins in the open. If the idea below interests you, we want you working on it with us — see CONTRIBUTING.md.
A photo uploaded straight from a phone or camera can contain:
- Precise GPS coordinates and altitude
- Capture timestamp
- Device make/model, serial number, software history
- Original filename and edit history
- Camera/lens settings (useful — but bundled in with the rest)
None of this is visible in the app UI. All of it can be read by anyone who downloads the file.
| Mode | Purpose | Behavior |
|---|---|---|
| 📷 Photography | Preserve what's useful for photographers | Keep ISO, aperture, shutter speed, focal length, camera/lens info. Strip GPS, altitude, device ID, timestamps. |
| 🔐 Privacy | Share with protection, not blindness | Encrypt sensitive metadata instead of deleting it — recoverable later by the owner, invisible to everyone else. |
| 🕳️ Zero-Metadata | Maximum minimization | Strip all embedded metadata as aggressively as possible before the file leaves the device. |
Long-term direction: camera integration, so the policy applies automatically at capture — open camera, pick a mode, shoot. No manual cleanup step people will skip.
See ARCHITECTURE.md for the technical design, metadata classification approach, and research questions this project is trying to answer.
Existing tools mostly do one thing: delete everything. MetaVeins treats metadata handling as a classification problem — what's useful, what's dangerous, and what should be a user decision — rather than a blunt strip operation. That's a harder, more interesting problem, and it's the reason this is worth building well rather than building fast.
This project is intentionally being built in the open from day one. Whether you're into privacy/security, mobile dev, image processing, or UX for non-technical safety tooling — there's a place to start. See CONTRIBUTING.md and docs/good-first-issues.md.
MIT — see LICENSE.