NotebookLM Research Skill

Create a NotebookLM notebook, optionally with URL and text sources. Args: title: Human-readable notebook title. sources: Optional list of URLs to ingest (web pages, PDFs, YouTube). text_sources: Optional list of raw text strings to add as sources. Returns: Notebook metadata including notebook id and sources added. Example: nlm_create_notebook( title="AI Safety Research", sources=["https://arxiv.org/abs/2401.00001"], text_sources=["Additional context about AI alignment..."] )

List all notebooks in the NotebookLM account. Returns: Dict with notebooks list, each containing id and title. Example: nlm_list()

Delete a notebook and all its sources. Irreversible. Args: notebook: Notebook ID or title to delete. Returns: Deletion status and deleted notebook ID. Example: nlm_delete(notebook="Old Research")

Add a source to an existing notebook. Provide exactly one of: url, text, or file_path. Args: notebook: Notebook ID or title. url: URL to add (web page, PDF, YouTube, etc.). text: Raw text content to add as a source. text_title: Title for the text source (default "Text Source"). file_path: Local file path to upload (PDF, TXT, etc.). Returns: Source metadata including source id and type. Example: nlm_add_source(notebook="My Research", url="https://example.com/paper.pdf") nlm_add_source(notebook="My Research", text="Key finding: ...", text_title="Notes")

Ask a question against a notebook's ingested sources. The answer is grounded in the notebook's content and includes citations. Args: notebook: Notebook ID or title to query. query: Natural-language question. Returns: Answer text with citations pointing to source passages. Example: nlm_ask(notebook="AI Safety Research", query="What are the main risks?")