notebooklm-connector

Integrates with Google NotebookLM to read sources, merge documents, and generate Studio outputs via MCP and browser automation.

5|Updated Jun 13, 2026
One-click install
npx skills add https://github.com/PatriotAi/ai-lab --skill notebooklm-connector-patriotai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: notebooklm-connector
Source: https://github.com/PatriotAi/ai-lab/tree/main/melania-skills-ecosystem/skills/notebooklm-connector
Command: npx skills add https://github.com/PatriotAi/ai-lab --skill notebooklm-connector-patriotai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? NotebookLM has no public API, so automating source extraction, merging, and content generation normally fails with 403 errors. This Skill provides a cascading set of access methods (Python MCP, Browserbase cloud browser, React Artifact OAuth, guided manual) so NotebookLM notebooks can be read, enriched, and turned into study or briefing materials from any device, including Android. ## Core Features & Use Cases - Multi-method access cascade: Tries notebooklm_* MCP tools (19 tools), then Browserbase cloud browser with screenshots, then a collaborative browser artifact, then guided manual steps — never stopping at an API error. - Source merging and import: Fetches multiple URLs in parallel, merges all notebook sources into one structured document, and prepares content for NotebookLM's 500,000-character Copied Text limit with cross-referenced splitting. - Studio feature generation: Guides generation of Audio Overview, Video Overview, Briefing Doc, Study Guide, FAQ, Timeline, Mind Map, Slide Deck, Infographic, Flashcards, Quizzes, and Deep Research with exact UI paths. - Use Case: A student shares a notebooklm.google.com link and asks to combine all sources into one document and generate a study guide; the Skill extracts content via MCP tools, produces a merged Markdown file, and gives exact Studio click paths plus five targeted Chat queries. ## Quick Start Ask the assistant to analyze the sources in your NotebookLM notebook URL and merge them into a single document with a recommended Studio feature.

Frequently Asked Questions about notebooklm-connector

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I access NotebookLM programmatically without an API?▼

NotebookLM has no public API, so access uses a cascade of methods: a Python MCP server with 19 tools via Google RPC, a Browserbase cloud browser MCP with screenshots, a React Artifact with Google OAuth, or guided manual copy-paste as a last resort.

How to merge all NotebookLM sources into one document?▼

The Skill collects content via MCP tools, Browserbase, or web_fetch, detects themes and removes duplicates, then outputs a structured Markdown document with a table of contents, per-source sections, synthesis, and a source index. Output can also be converted to .docx or .pptx.

Can I use NotebookLM automation on Android?▼

Yes. The recommended Android path uses the Browserbase cloud browser MCP with the Cookie Editor app to export Google session cookies, which are injected via browserbase_authenticate. Every action returns a screenshot, and no terminal or npm is required on the phone.

What NotebookLM Studio features can be generated?▼

Supported Studio outputs include Audio Overview, Video Overview, Briefing Doc, Study Guide, FAQ, Timeline, Mind Map, Slide Deck, Infographic, Flashcards, Quizzes, and Deep Research. The Skill provides exact UI navigation paths and notes limits like the free-tier Slide Deck watermark.

Why does NotebookLM return 403 errors and how is it handled?▼

Google blocks non-browser requests with 403 because NotebookLM requires an authenticated session. The Skill silently cascades to the next access method — MCP tools, cloud browser, or OAuth artifact — instead of stopping to explain the limitation.

What is the size limit for adding text sources to NotebookLM?▼

A Copied Text source accepts roughly 500,000 characters. Content exceeding this limit must be split into multiple parts with cross-references rather than pasted as a single block.