Phone and browser capture

A box is a directory of your data, kept under version control with a full history of changes (using git); a card is a markdown file with a structured header; the agent is the coding agent (Claude Code or Codex) that operates the box. Capture is how content from your phone or browser gets into the box.

What it does for you

• Pairs your iPhone to a box by scanning a QR code or opening a link, so captures are attributed to you without signing in each time.
• Lets you record voice, take photos, or upload files from the native app, landing in the box's web chat as a normal message.
• Keeps an interrupted or backgrounded capture uploading instead of losing it, and tells you plainly if a recording got cut off.
• Groups everything from one recording session into a single timeline you can read back, with silences and photos in place.
• Lets you see and revoke paired devices, and lock a paired box behind Face ID or your device passcode.

What it needs

An iPhone (iOS 17 or later) for the native app, or Chrome for the extension. Both need a box already reachable from your device. See ../install/index.md.

How it works, briefly

The iOS app is a thin native shell around the same web chat you use in a browser; it adds native pairing, recording, and speech input, delivered into the chat over a bridge. A finished capture becomes a capture-session card with typed cards for what it contains (image, audio, file). Capture runs when you use it, not on a schedule. Clipping web pages is a separate capability: web-clipping.md.

Limits

Owner-gated surfaces like device pairing require signing in as the box owner; a shared or guest device cannot pair itself. The documentation does not describe an Android app.

Go deeper

../reference/cards/capture-session.md, ../reference/cards/image.md, ../reference/cards/audio.md, ../reference/cards/upload-batch.md, ../contracts/mobile-contract.md