List the user's NotebookLM notebooks, owned and shared. Returns each notebook's id, title, emoji, sources (id, title, type), ownership and timestamps, plus a count. Start here to find a notebook_id by title; once you have the id, notebook_get reads one notebook. There is no paging: if count equals max_results, call again with a higher max_results.

Create an empty NotebookLM notebook. Returns the new notebook; keep its id for every later call. Each call creates another notebook, so first look for one with this title in notebook_list. Then fill it: source_add_url, source_add_text or source_add_drive for sources you have, research_start to find some.

Get one notebook's details. Returns its title, emoji, ownership, timestamps and sources, each with id, title and type (web_page, youtube, pasted_text, pdf, google_docs, ...). Use it to find source IDs for notebook_query and the Studio tools, or to check that a source you just added has appeared.

Ask NotebookLM for an AI-generated overview of the whole notebook. Returns a short summary of what its sources cover and the questions NotebookLM suggests asking about them. Use it to get oriented in an unfamiliar notebook; ask specific questions with notebook_query and list the sources with notebook_get.

Change a notebook's title; its sources and chat stay as they are. Returns a confirmation message with the new title. Find the notebook_id with notebook_list first.

Permanently delete a notebook and everything in it. With confirm=false it does nothing and returns pending_confirmation; with confirm=true it deletes and returns a confirmation message. This cannot be undone: name the notebook to the user and get explicit approval first.

Get one source's metadata. Returns its id, title, type (web_page, youtube, pasted_text, pdf, google_docs, ...) and content_length, the number of characters NotebookLM extracted from it. The text itself is left out because it can be very long; fetch it with source_get_content.

Get the full text NotebookLM extracted from a source, with tables flattened to "cell | cell" rows. Returns { text }, which can be long: a long Wikipedia article came to about 100,000 characters. Prefer notebook_query to answer questions; use this when the user needs the verbatim text.

Add a web page or a YouTube video to a notebook by URL. NotebookLM fetches the page itself, so it must be publicly reachable. Returns the new source (id, title, type web_page or youtube); the title can show the URL until processing finishes. If NotebookLM cannot fetch the page the call fails: check notebook_get before retrying.

Add text as a source: pass it in content, or point file_path at a local file (content wins if both are set). A file_path outside the server's working directory and the OS temp directory is refused; read such a file yourself and pass content. Returns the new source (id, title, type pasted_text). For a web page or YouTube video use source_add_url; for a Drive file, source_add_drive.

Add a Google Drive file as a source, by its Drive file ID and MIME type; the signed-in Google account must be able to open the file. Returns the new source (id, title, type). The source is a snapshot: when the Drive file changes, refresh it with source_sync_drive. For a public web page use source_add_url instead.

Re-import Google Drive sources (added with source_add_drive) so the notebook uses the latest version of each file, replacing the stored snapshot. With confirm=false it does nothing and returns pending_confirmation; with confirm=true it syncs and returns how many sources were synced.

Permanently remove one source from a notebook. With confirm=false it does nothing and returns pending_confirmation; with confirm=true it deletes and returns a confirmation message. This cannot be undone: name the source to the user and get explicit approval first.

Start generating an Audio Overview: a podcast-style conversation between two AI hosts about the notebook's sources. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. Once completed, studio_status gives an audio download_url.

Start generating a Video Overview: a narrated video with visuals that explains the notebook's sources. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. Once completed, studio_status gives a video download_url.

Start generating an infographic: a single image that summarises the notebook's sources visually. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. Once completed, studio_status gives an image download_url.

Start generating a slide deck from the notebook's sources. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. Once completed, studio_status gives a download_url for the deck as a PDF.

Start generating a written report from the notebook's sources: a Briefing Doc, Study Guide, Blog Post, or your own brief with report_format "Create Your Own" and custom_prompt. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. This artifact has no download file: the user opens it in the notebook in NotebookLM.

Start generating a set of study flashcards from the notebook's sources. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. This artifact has no download file: the user opens it in the notebook in NotebookLM.

Start generating a multiple-choice quiz from the notebook's sources. Uses the user's NotebookLM quota and takes several minutes, so agree on the options with the user first: with confirm=false it does nothing and returns pending_confirmation. With confirm=true it returns the artifact_id; poll studio_status until that artifact is completed. This artifact has no download file: the user opens it in the notebook in NotebookLM.

List the notebook's Studio artifacts. Returns each one's id, type (audio, video, report, flashcards, quiz, infographic, slide_deck), status (pending, generating, completed or failed) and download_url: set for completed audio, video, infographics and slide decks (PDF), null otherwise. It lists older artifacts too: after a *_create call, poll every 30-60 seconds and find yours by its artifact_id.

Re-read the NotebookLM page with the saved cookies to refresh the CSRF token and session ID. Returns a confirmation, or an authentication error if the Google session is signed out; then ask the user to run npx -y @m4ykeldev/notebooklm-mcp auth in a terminal. Call it once after an auth error, then retry.

Manual fallback: save NotebookLM cookies the user pasted, plus an optional CSRF token and session ID, to ~/.notebooklm-mcp/auth.json, replacing what is stored there. Returns a confirmation; the next call uses the new tokens. Prefer the CLI auth command. Use this only when the user hands you cookies, and never echo them back.

Ask NotebookLM a question that it answers only from the notebook's sources (all of them, or source_ids). Returns the answer and a conversation_id. The answer text keeps NotebookLM's bracketed citation numbers such as [1], but not which passage each number points to. Pass conversation_id back to ask a follow-up; this server remembers the conversation's history while it keeps running.

Start a NotebookLM research task that searches the web, or the user's Google Drive, for sources on a topic. Deep mode is more thorough and slower than the default fast mode. Returns a task_id: poll research_status with it, then import the results you want with research_import.

Check research tasks in a notebook: one task, or all of them. Returns each task's status (in_progress, completed, imported, import_failed, or unknown for a state this server doesn't recognise), query, summary and found sources (url, title, description, type). Poll until the status is no longer in_progress; a source's position in this list is the index research_import takes.

Add results of a finished research task to the notebook as sources. Only web results can be imported: without source_indices it imports every web result and skips Drive ones. NotebookLM fetches each URL and skips any it can't reach, so it returns "Imported X of Y research sources". If none could be fetched the call fails and the task shows import_failed; a retry may succeed.