Save what matters. Find it by what you remember. Keep the context that makes it yours. seenX starts with screenshots and images, then keeps your own thoughts connected to the original material.

One memory, two ways in
Search for yourself in a lightweight command bar, or authorize a compatible assistant to retrieve relevant images and notes through the same local engine. Assistant access is optional. The library should be useful on its own.
Keep the reason, too
An image can show a checkout flow. A note can say “keep the spacing, change the colors.” That distinction matters: your intention is explicit user context, not something a model should guess.
How notes and saved context fit together →
A lightweight first release
The first-release plan centers on a global search shortcut, a small capture HUD for optional notes, selected image sources, local retrieval, and an opt-in MCP bridge. A heavy organizational dashboard is not a prerequisite. PDF and bookmark support can follow.
What works today
The website preview includes a pretend screenshot flow, image search, text highlights, pinned notes, and JSON context export. Notes, selected text, and their image coordinates are saved in this browser. It does not access your Mac, run semantic models, capture application activity, or connect to an assistant. Native capture and retrieval integration remain in development.