jacob-bd/gemini-notebook-mcp-cli

Programmatic access to Gemini Notebook - via command-line interface (CLI), Model Context Protocol (MCP) server, and AI agent skills.

Gemini Notebook (formerly Google NotebookLM) CLI & MCP Server

What it is – A Python package that lets you control Google’s Gemini Notebook (the service formerly known as Notebook LM) from the command line or via a Model Context Protocol (MCP) server. The MCP server can be hooked up to AI‑assistant tools such as Claude, Cursor, Gemini‑CLI, GitHub Copilot, etc., so those agents can read, write and generate content in your Gemini Notebook automatically.

Key capabilities

Area CLI command (example) MCP tool name
List notebooks nlm notebook list notebook_list
Create a notebook nlm notebook create "Research Project" notebook_create
Add sources (URL, text, Drive file, upload) nlm source add <nb> --url "https://…" source_add
Query a notebook (adds a chat entry in the web UI) nlm notebook query "Summarize key points" notebook_query
Manage chat sessions nlm chats list / nlm chats get <id> chat_list, chat_get
Generate Studio content (audio, video, slides, infographics, etc.) nlm studio create <nb> audio studio_create
Revise slide decks nlm slides revise <nb> studio_revise
Download artifacts nlm download audio <nb> <artifact-id> download_artifact
Bulk download all artifacts nlm download all <nb> -d ./exports download_all_artifacts
Sync Drive‑based sources nlm source sync source_sync_drive
Batch operations, cross‑notebook queries, pipelines, tagging, etc. nlm batch …, nlm cross query …, nlm pipeline run …, nlm tag add … batch, cross_notebook_query, pipeline, tag
Share notebooks (public link or invite) nlm share public <nb> notebook_share_*
Configure AI tools (Claude, Cursor, Gemini‑CLI, etc.) nlm setup add claude-code –
Install “skills” that teach the assistant how to use the MCP nlm skill install cline –
Diagnose problems nlm doctor –

How you use it

  1. Install – pip install notebooklm-mcp-cli (or the recommended uv tool install notebooklm-mcp-cli).
  2. Authenticate – nlm login opens a browser, extracts Google cookies, and stores them locally. Profiles let you keep separate Google accounts.
  3. Run the CLI – nlm … lets you script notebook creation, source addition, content generation, downloading, sharing, etc.
  4. Expose the MCP server – notebooklm-mcp starts a local HTTP server that AI assistants can call. With nlm setup add <tool> the package writes the appropriate JSON configuration for tools like Claude Desktop, Cursor, GitHub Copilot, etc., so those agents can invoke the 43 built‑in MCP tools.
  5. Talk to your assistant – Once the MCP is registered, you can simply say things like “Create a notebook about quantum computing and generate a podcast” and the assistant will call the appropriate MCP endpoints behind the scenes.

Installation & upgrade

  • Preferred: uv tool install notebooklm-mcp-cli (or uvx for a one‑off run).
  • Also available via pip / pipx.
  • Upgrading follows the same command (uv tool upgrade … or pip install --upgrade …).
  • Legacy packages (notebooklm-cli, notebooklm-mcp-server) must be removed before installing the unified package.

Safety note – The tool talks to undocumented internal Google APIs and requires you to export cookies from your browser. Those APIs can change without notice, so the package is intended for personal/experimental use only.

Who it’s for – Developers, researchers, or power users who want programmatic or AI‑agent‑driven access to Gemini Notebook for tasks such as automated research aggregation, content generation (audio, video, slide decks, infographics), bulk export, or sharing notebooks without manually using the web UI.

Quick start example

# Install
uv tool install notebooklm-mcp-cli

# Log in (opens browser, extracts cookies)
nlm login

# Create a notebook and add a source
nlm notebook create "AI Strategy" 
nlm source add "AI Strategy" --url "https://arxiv.org/abs/2302.12345"

# Ask the notebook to summarize
nlm notebook query "Summarize the main findings"

# Generate a podcast and download it
nlm studio create "AI Strategy" audio
nlm download all "AI Strategy" -d ./exports

MCP registration example (for Claude Desktop)

nlm setup add claude-code   # writes the needed JSON into Claude’s config

Now Claude can run commands like @gemini-notebook-mcp notebook_create "My Notebook".

Resources

  • Full docs: docs/GETTING_STARTED.md, docs/CLI_GUIDE.md, docs/MCP_GUIDE.md
  • Demo videos linked in the README showcase both CLI usage and MCP integration with various assistants.
  • The package is published on PyPI as notebooklm-mcp-cli.

Related

  • Project
    teng-lin/notebooklm-py

    An unofficial Python API and CLI for Google's NotebookLM (Gemini Notebook), enabling programmatic research automation, multimedia content generation, and grounded memory for AI agents.

  • Project
    claude-world/notebooklm-skill

    An automation tool and MCP server for Google NotebookLM that enables creating notebooks, performing grounded research, and generating artifacts via CLI or AI agents.

  • Project
    arabold/docs-mcp-server

    Grounded Docs MCP Server is a local, version‑aware documentation indexer that scrapes official docs (web, GitHub, npm, PyPI, local files), converts them to searchable text, and optionally adds vector embeddings. It exposes a simple MCP (SSE) API so AI coding assistants (Claude, Cline, Copilot, Gemini‑CLI, etc.) can retrieve exact documentation passages, reducing hallucinations while keeping data private.

  • Project
    matlab/matlab-mcp-server

    MATLAB MCP Server is a Go‑based service that lets AI assistants (Claude, Copilot, Codex, etc.) start MATLAB, run scripts, perform static code checks, and query toolbox info via the Model Context Protocol. It ships pre‑built binaries for all major OSes, supports custom tool extensions, and can attach to existing MATLAB sessions. Intended for developers who want to embed real MATLAB computation inside LLM‑driven workflows, it requires a valid MathWorks license and offers configurable startup options, logging, and optional telemetry.

  • Project
    WJZ-P/gemini-skill

    A tool that automates the Gemini web interface via CDP and MCP, enabling AI agents to perform text conversations, image generation, and image extraction.