Skip to main content

Your First Capture

Once the extension is installed and your BYOK keys are configured, you are ready to capture your first piece of knowledge.

Capture an Article​

  1. Navigate to any article or blog post in Chrome.
  2. Click the Nexus icon in the toolbar.
  3. Click the "Capture" button in the popup.
  4. The extension will:
    • Scrape the main article text from the page.
    • Call your LLM API to generate a summary and extract entities (people, tools, concepts).
    • Save the node to your Supabase database.
    • Generate a text embedding and auto-link it to related existing nodes.

The popup displays a progress indicator with the current step (e.g., "Extracting content…", "Processing with AI…", "Saving to database…").

Capture a YouTube Video​

  1. Navigate to any YouTube video that has captions enabled.
  2. Click the Nexus icon and then "Capture".
  3. The extension extracts the video transcript from the page, then processes it the same way as an article.
No Transcript Available

If a YouTube video has no captions, the extension will display an error: "No transcript available to capture." This is expected — the video cannot be processed without text content.

View Your Capture in the Dashboard​

After a successful capture:

  1. Open the web dashboard.
  2. Your new node appears at the top of the Feed list on the left panel.
  3. If it shares semantic similarity with any existing nodes (cosine similarity > 0.8), edges are automatically created and visible in the Graph tab.

The Capture Counter​

The extension popup header shows a daily capture count badge (e.g., "3 today"). This resets at midnight local time.

Capture History​

The extension popup also shows a short list of your recent captures with their titles. Clicking on a history item does not navigate anywhere — it's a quick reference for what you've captured.

Next Steps​