Use case

Save links from screenshots

Recover URLs from iPhone screenshots and save them as openable resource cards with titles, tags, and metadata.

Last updated: 2026-05-09

gh repo clone ReScienceLab/SnapAction
Short answer

SnapAction extracts links from iPhone screenshots, recovers missing canonical URLs when only a title or name is visible, and stores the result as a typed resource card you can open later.

A screenshot of a tweet, chat, or article often shows a title or mention without the actual URL. The context is clear when you take it, but a week later the image is just another screenshot in Photos.

Copying links manually is slow and easy to forget. The result is a growing library of screenshots you meant to revisit but never did.

SnapAction uses a Convex-backed OpenRouter agent to analyze the screenshot. If a URL is visible, it is extracted directly. If only a title, repo name, product, or place is visible, SnapAction can use Serper search and URL verification to recover the likely canonical link before saving the card.

The saved card includes the title, URL, type, tags, description, and an Open action so the link is usable immediately.

Link recovery works best when the screenshot contains a clear primary resource: a GitHub repo page, an article title, a product name, a place name, or a video title. Screenshots with ambiguous titles, heavy cropping, or multiple unrelated items may produce less useful results.

The link is stored in a local SwiftData resource card alongside tags, metadata, and the screenshot reference. You can open it directly from SnapAction, mark it as read or favorite, and find it again through search or the Rewind daily digest.

FAQ

Can SnapAction find links in screenshots without a URL?

Yes. When a title, repo name, product, or place is visible, SnapAction can use Serper search and URL verification to recover a likely canonical link.

What if the screenshot shows multiple links?

SnapAction focuses on the primary 1-3 resources in a screenshot or consecutive screenshot group. It intentionally extracts the main resource rather than every incidental piece of text.

Next step

Turn iPhone screenshots into action-ready resource cards with AI.

View on GitHub