notebooklm-research

Runs cited research and synthesis over NotebookLM notebooks via the notebooklm MCP server.

Updated Jun 28, 2026
One-click install
npx skills add https://github.com/JaviMontano/claude-plugins --skill notebooklm-research-javimontano
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: notebooklm-research
Source: https://github.com/JaviMontano/claude-plugins/tree/main/plugins/claude-native-toolkit/skills/notebooklm-research
Command: npx skills add https://github.com/JaviMontano/claude-plugins --skill notebooklm-research-javimontano

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) components.

What problem does it solve? Turning heterogeneous sources (URLs, documents, text, Drive files) into queryable, citation-backed research context is manual and error-prone. This Skill orchestrates the NotebookLM MCP server so every synthesized claim stays linked to its source, avoiding unsourced summaries. ## Core Features & Use Cases - Notebook and source management: Create or select notebooks and add heterogeneous sources (URL, text, Drive, file), then verify their content. - Cited querying and research: Run single-notebook queries, federated cross-notebook queries, or multi-document research jobs that return cited syntheses. - Studio artifacts: Optionally generate audio, slides, or infographics from notebook sources with bounded polling for long-running jobs. - Use Case: Ask to research "EU AI Act obligations for SaaS" — the Skill creates a notebook, adds sources, runs the research pipeline, and returns a synthesis where every claim is attributed to a source. ## Quick Start Use NotebookLM to research a topic by building a notebook from my sources and giving me a cited synthesis of what they agree on.

Frequently Asked Questions about notebooklm-research

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

FAQPage Schema
How do I research a topic with NotebookLM and get cited sources?▼

Create or select a notebook, add your sources (URLs, text, Drive, or files), then run a notebook query or start a research job. The synthesis returned keeps every claim linked to its source, and unsourced gaps are flagged as coverage_gap.

How do I query across multiple NotebookLM notebooks at once?▼

Use the cross_notebook_query tool for federated search across several notebooks. List notebooks first with notebook_list to confirm targets exist, since running a cross-notebook query without target notebooks returns zero results.

NotebookLM research vs web scraping: which should I use?▼

Use NotebookLM research when you need multi-source synthesis with citations from documents you provide. For live scraping of a single web page, use a web-research approach such as firecrawl instead, since this Skill is a harness over the notebooklm MCP server.

Why does the NotebookLM MCP server not respond in my session?▼

The server must be probed in a live session before use; run notebook_list and check for status:success. If authentication fails, run nlm login in the terminal. In headless runs the server may be absent entirely.

What happens when a NotebookLM research or studio job never completes?▼

Long-running jobs like research_status and studio_status are polled with a fixed maximum number of attempts. If the job never completes due to quota or timeout, the Skill reports a coverage_gap instead of blocking indefinitely.