Notebooklm Research

NotebookLM Research Automation

Turn Claude Code into a research agent by automating Google NotebookLM. NotebookLM has no public API. This skill uses Playwright to control a real Chrome browser via CDP (Chrome DevTools Protocol). All commands output JSON to stdout.

Prerequisites

One-time setup (2 minutes):

pip install playwright && playwright install chromium
python scripts/setup_chrome.py
# Log in to Google in the Chrome window that opens
python scripts/notebooklm_client.py status  # Verify: "authenticated": true

See references/authentication.md for detailed setup and troubleshooting.

Session Management

BAD: Launching Chromium directly from Playwright. Gets detected as a bot, cannot authenticate with Google.

browser = await playwright.chromium.launch()
page = await browser.new_page()
await page.goto("https://notebooklm.google.com")  # Blocked or login fails

GOOD: Connecting to a real Chrome instance via CDP with a persistent profile.

browser = await playwright.chromium.connect_over_cdp("http://localhost:9222")
context = browser.contexts[0]  # Reuse authenticated session
page = context.pages[0]        # Already on NotebookLM

Interaction Speed

BAD: Instant typing and clicks. Triggers anti-bot detection, actions may be ignored.

await page.fill("textarea", "full text instantly")
await page.click("button")

GOOD: Human-like delays. Random 25-75ms per character, 100-300ms pre-click pause.

for char in question:
    await element.type(char, delay=random.uniform(25, 75))
await asyncio.sleep(random.uniform(0.1, 0.3))
await page.click("button")

Command Reference

Always verify connection first with status. All commands return JSON.

Artifact types: slides, infographic, quiz, flashcards, report, table, mindmap, video

Research Workflow

Follow this sequence for reliable results:

1. Check connection: status command. If not connected, run setup_chrome.py.
2. Create notebook: create "Research Topic Name"
3. Add sources: Run add-source for each URL, file, YouTube link, or text block. Add 2-3+ sources for best results.
4. Wait for processing: Sources need 10-60 seconds to index. The script waits automatically, but for large PDFs allow extra time.
5. Query sources: query --notebook "Topic" --question "Your question". Returns answer with citations.
6. Generate deliverables: generate --notebook "Topic" --type slides. Output saved to ./notebooklm-output/.

Output Format

Success:

{"status": "success", "answer": "...", "citations": ["Source 1, p.3", "Source 2, sec.4"]}

Error:

{"status": "error", "error": "Chrome not reachable", "suggestion": "Run: python setup_chrome.py"}

Generated files are saved to ./notebooklm-output/ (override with NOTEBOOKLM_OUTPUT_DIR env var).

Troubleshooting

See references/deliverables.md for artifact types, generation times, and examples. See references/authentication.md for Chrome setup, session management, and security.