메뉴
Google NotebookLM integration - create notebooks, add sources, generate AI content (podcasts, quizzes, mind maps, briefings, slides), query notebooks, run research agents, and manage research workflows
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Ask which skill or flow fits your situation. A router over the user-invoked skills in this repo.
Stages all changes and generates a lean Conventional Commits message via caveman-commit — user reviews and commits manually. Triggers on any commit intent: "commit", "stage my changes", "ready to push", "let's commit", "make a commit", "commit this", "push changes", "squash and commit", or any variation. Use whenever the user signals they're done with changes and ready to commit, even if they don't mention "message" or "caveman". Does NOT run git commit — outputs a ready-to-run command for the user to execute after review.
Autonomous ML research - agent modifies GPT training code, runs 5-min experiments, keeps improvements
Always use browser-harness for any web interaction: automation, scraping, testing, or site/app work.
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
AI spend analytics — see where your token budget goes by task, model, tool, and project across 31 AI coding tools including Claude Code. Use when the user asks about AI costs, token usage, spend breakdown, or wants to optimize their AI budget.
| name | notebooklm |
| description | Google NotebookLM integration - create notebooks, add sources, generate AI content (podcasts, quizzes, mind maps, briefings, slides), query notebooks, run research agents, and manage research workflows |
Programmatic access to Google NotebookLM via CLI (notebooklm) and MCP server (notebooklm-mcp).
Source: https://github.com/teng-lin/notebooklm-py
Use notebooklm CLI commands via Bash tool for all operations.
# Auth
notebooklm login # browser sign-in (one-time)
notebooklm auth check --test --json # verify auth
notebooklm doctor --fix # diagnose + auto-fix issues
# Notebooks
notebooklm list
notebooklm create "Research Project" --use # create + set active
notebooklm use <id> # set active notebook
notebooklm rename "New Title"
notebooklm summary --topics
# Sources
notebooklm source add --url "https://example.com"
notebooklm source add --file /path/to/doc.pdf
notebooklm source add --text "Pasted content here"
notebooklm source add --youtube "https://youtube.com/watch?v=..."
notebooklm source list
notebooklm source delete <source-id>
# Chat
notebooklm chat ask "What are the key findings?"
notebooklm chat ask "Summarize this" --save-to-notes # persist answer as note
# Content generation
notebooklm audio create # podcast
notebooklm audio status <task-id>
notebooklm audio download <task-id> --output podcast.mp3
notebooklm video create
notebooklm video status <task-id>
notebooklm video download <task-id> --output video.mp4
notebooklm quiz create --difficulty medium
notebooklm quiz download <task-id> --output quiz.json --format json
notebooklm mind-map create
notebooklm mind-map download <task-id> --output mindmap.json
notebooklm briefing create
notebooklm flashcards create
notebooklm infographic create --orientation landscape
notebooklm slides create
# Notes
notebooklm note list
notebooklm note create "My insight"
notebooklm note delete <note-id>
# Research agents
notebooklm research add-web "enterprise AI adoption 2025"
notebooklm research add-web "cloud costs" --mode deep
notebooklm research add-drive "product roadmap"
# Sharing
notebooklm share public # make public
notebooklm share invite user@example.com --role viewer
# Profiles (multiple Google accounts)
notebooklm profile create work
notebooklm profile switch work
notebooklm profile list
# MCP auto-config (run once)
notebooklm mcp install claude-code
Only use when CLI cannot handle the task. Common MCP tools via notebooklm-mcp:
notebook_list, notebook_create, notebook_delete, notebook_querysource_add, source_list, source_delete, source_sync_drivestudio_create (audio, video, quiz, mind_map, briefing, flashcards, infographic, slides)download_artifact, export_artifactnotebook_share_public, notebook_share_inviteresearch_start (web + Drive discovery)note, label, tag, batch, pipelineFirst use:
uv tool install "notebooklm-py[browser]"
notebooklm login # browser opens, Chromium downloads ~170MB first time
notebooklm auth check --test --json # expect {"status": "ok"}
Headless (CI/server):
notebooklm login --master-token --account you@gmail.com
Auth expires: notebooklm auth refresh or notebooklm login again.
notebooklm create "Cloud Trends 2025" --use
notebooklm research add-web "cloud marketplace growth 2025" --mode deep
notebooklm chat ask "What are the top 3 trends?"
notebooklm audio create
notebooklm audio status <task-id> # poll until done
notebooklm audio download <task-id> --output podcast.mp3
# All generation is async — create → status → download
notebooklm audio create && notebooklm audio status <id> && notebooklm audio download <id>
notebooklm quiz create --difficulty hard && ...
notebooklm mind-map create && ...
notebooklm slides create && ...
notebooklm research add-drive "product roadmap" --max-results 10
notebooklm source add --drive "<file-id>"
| Task | Use |
|---|---|
| All single operations | CLI |
| Batch ops across many notebooks | MCP: batch tool |
| Pipeline (create→add→generate) | MCP: pipeline tool |
| Cross-notebook queries | MCP: cross_notebook_query |
| Structured output needed in-context | MCP |
Default: CLI for everything. MCP is backup only.
Already configured. On fresh clones, setup.sh installs via:
uv tool install "notebooklm-py[browser]"
Upgrade: uv tool upgrade notebooklm-py
notebooklm doctor --fix # auto-detect + fix common issues
notebooklm auth check --test # test auth
notebooklm --version # check version