How to Tag Amazon Synthetic-Performer XMP Metadata for Free

If your listing images contain photorealistic AI-generated people, they may need to carry specific metadata identifying that before you upload them. Doing this by hand, file by file, across a real catalog is tedious and easy to get wrong. This tool turns it into a review queue: it scans a folder, shows you which images are untagged, and lets you approve or skip each one — nothing is tagged automatically without you looking at it.

How the workflow works

1. Organize your images under Media Library\Seller-SKU\... folders.
2. Open the app, point it at that Media Library folder, and click Scan. Nothing runs in the background — it only scans when you ask.
3. Review the untagged images for one Seller SKU at a time. Click an image to select or deselect it, then choose Add tag & verify or No tag needed.
4. Upload the prepared files to Seller Central yourself — the app never logs in to Amazon or claims an upload succeeded.

Use Select visible for the current SKU, or Select all media across the whole queue. In Duplicates & visual variants, mark a false match as Not a duplicate to remember that decision by content fingerprint — exact byte-for-byte duplicates can't be dismissed the same way.

What "verified locally" actually means

Before the first change to a file, the app creates a private local safety backup. It then appends the keyword — without touching any other subject tags — and reads dc:subject again. A file only counts as successful if that exact keyword is present after the re-read. Safety backups are eligible for manual cleanup after 30 days and are never deleted automatically.

Privacy and safety

• Your media, review decisions, backups, and app settings stay on your computer — no cloud or network AI integrations.
• The app window itself has no direct filesystem access outside the folder you selected that session.
• Images with existing valid tags are treated as already approved and skip the review queue.
• File deletion goes through the Windows Recycle Bin or macOS Trash only.

Platforms and languages

Windows and macOS, with English, Simplified Chinese, and Traditional Chinese interfaces, light/dark/auto theme, and a five-step quick-start tutorial for new users. Supported image formats: JPEG/JPG, PNG, TIFF/TIF, WebP. Videos show up under "Other files" for a manual watch-and-tag workflow.

The Mac build isn't Apple-notarized yet, so macOS will warn on first open — download the right .dmg for your chip (arm64 for Apple Silicon, x64 for Intel) from the Releases page, then Control-click the app in Applications and choose Open to approve it once.

FAQ