PDF Summariser — Browser Extension + NotebookLM API

Chrome extension + Python backend to summarise PDFs with NotebookLM.

What It Does

• Detects PDF tabs in Chrome
• Lets user generate:
  • Report (PDF)
  • Presentation (PPTX)
  • Podcast (MP4)
  • Quiz (PDF)
  • Chat with PDF
  • Export report to Google Docs

Architecture — Per-User Auth

Each user authenticates with their own Google account. No shared session on the server.

1. User installs the extension
2. First use: clicks "Login" in popup → opens NotebookLM in a new tab → signs into Google
3. Extension reads their Google cookies via chrome.cookies API
4. Every API request includes cookies as X-NLM-Auth header
5. Backend creates a per-user NotebookLM client, cached for 1 hour

You (the developer) deploy the backend once. No session tokens to refresh. Ever.

Project Structure

• extension/ - Chrome MV3 extension
• summariser_api.py - FastAPI backend
• API/notebooklm-py/ - NotebookLM client library
• Dockerfile - container deployment
• privacy-policy.html - store privacy policy
• store-listing.md - Chrome Web Store listing copy

Local Development

1. Create venv and install deps
2. Install local NotebookLM client package
3. Start API:

python -m uvicorn summariser_api:app --reload --port 8000

4. Load extension/ as unpacked extension in chrome://extensions
5. Click extension icon → Login → sign into NotebookLM → done

The server no longer requires NOTEBOOKLM_AUTH_JSON — auth comes from each user's browser.

Cloud Deployment

See RENDER_DEPLOYMENT.md for full Render deployment steps.