#783 · Primary category: AI Agents & Automation

notebooklm-mcp

MCP server for NotebookLM - Let your AI agents (Claude Code, Codex) research documentation directly with grounded, citation-backed answers from Gemini. Persistent auth, library management, cross-client sharing. Zero hallucinations, just your knowledge base.

Project last updated:05/01/26

GitHub Stars

3.4K

Forks

485

Contributors

2

License

MIT

Why we included this project

Teams that keep research material inside Google NotebookLM often find their coding agents live in a separate world. This MCP server closes that gap by exposing NotebookLM as a tool that Claude Code, Codex, Cursor, or any Model Context Protocol client can call directly. It drives a real Chrome browser through Patchright with a persistent profile, so setup takes one interactive login and every run after that is headless. Agents can chat against a notebook, add new sources, generate audio overviews, and read DOM-level citations, which lets each answer point back to the specific source it came from. That grounding is what makes the results usable for documentation and internal research instead of unverified model output. Both stdio and Streamable-HTTP transports are supported, so tools that only speak HTTP can reach the same notebooks too.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category