Capturing Content

The extension's primary function is to capture web content and process it with AI.

Supported Content Types

Content Type	Source	Method
Articles	Any web page with <article> element	DOM text extraction
YouTube Videos	YouTube.com pages	Transcript extraction from captions

Capture Status States

The popup displays one of these status states during capture:

Status	Description
idle	Ready to capture.
extracting	Extracting text content from the page.
processing	Calling the LLM API to generate summary and entities.
done	Capture complete. Node saved to database.
error	Something went wrong. Error message is displayed.
jwt_expired	The Supabase session has expired. Click Sign In in the popup.

AI Processing Output

The LLM call produces a structured JSON response:

{
  "summary": "A concise 2-3 sentence summary of the content.",
  "entities": [
    { "name": "React", "type": "tool" },
    { "name": "Dan Abramov", "type": "person" },
    { "name": "Component Architecture", "type": "concept" }
  ]
}

This data is saved to the nodes table (summary) and entities table.

Daily Capture Counter

The popup header displays how many items you have captured today. The counter:

• Resets at midnight local time.
• Is stored with a date key in local storage.
• Is shown as a badge on the extension icon (if supported by the browser).

Capture History

The popup shows a scrollable list of your recent captures. Each entry displays:

• Page title.
• Source URL (truncated).
• Capture timestamp.

This is a local-only display from storage — it does not load from the database.

Authentication Status

The extension monitors your Supabase session and displays your authentication status in the popup. If you are not authenticated, you will see a Sign In button. Once signed in, your session automatically refreshes.

When signed out, captures will fail with a "Not authenticated" error. Click Sign In from the extension popup to reconnect to your dashboard session.