Quick Capture
Quick Capture saves the current browser tab to your LoomBrain knowledge graph.
How it works
Section titled “How it works”- Navigate to a page you want to save
- Click the LoomBrain extension icon
- Optionally type a “why” note — explain why you’re saving this
- Optionally select text on the page before opening the popup
- Click Capture Tab
The extension extracts the page content, detects its type, and sends it to the ingestion pipeline. Within about a minute, it appears as a searchable node in your graph.
Content type detection
Section titled “Content type detection”The extension automatically detects content type from the URL:
| URL pattern | Detected type |
|---|---|
twitter.com/*/status/* or x.com/*/status/* | Tweet |
youtube.com/watch?v=* or youtu.be/* | Video |
github.com/{owner}/{repo} (2 path segments) | Repository |
| Everything else | Article |
Each type uses a specialized extraction strategy:
- Tweet: Extracts the tweet element HTML
- Video: Sends URL only — the server fetches transcript and metadata
- Repository: Extracts README content plus repo description and topics
- Article: Extracts the full page HTML, sanitized
Selected text
Section titled “Selected text”If you select text on the page before clicking the extension, it’s included with the capture as selected_text. This is useful for highlighting a specific passage that prompted you to save the page.
Fallback behavior
Section titled “Fallback behavior”If content extraction fails or the page HTML exceeds 5MB, the extension falls back to URL-only capture. The server then fetches and processes the content on its own.
The extraction has a 10-second timeout. If the page takes longer to extract (heavy SPAs), it falls back to URL-only.