Back to directory

notebooklm-py

An unofficial Python API and agent skill for Google's Gemini Notebook (formerly NotebookLM), giving programmatic access to research automation, content generation and file exports from Python, a CLI, or AI agents such as Claude Code and Codex.

notebooklm-py

notebooklm-py is an unofficial Python library and agent skill for Google's Gemini Notebook, the product formerly known as NotebookLM. It exposes capabilities the web interface does not, through Python, a command-line tool, and AI agents such as Claude Code, Codex and OpenClaw.

It supports bulk-importing sources such as URLs, PDFs, YouTube videos and Google Drive files, running web and Drive research queries with automatic import, and extracting insights programmatically to build repeatable research pipelines. It can also generate audio overviews, videos, slide decks, quizzes, flashcards, infographics, data tables, mind maps and study guides, and download the results as MP3, MP4, PDF, PNG, CSV, JSON or Markdown files, or export them to Google Docs and Sheets.

The library relies on undocumented Google APIs, is not affiliated with Google, and may break if those endpoints change; its authors recommend it for prototypes, research and personal projects rather than production use.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.