notebooklm
/notebooklm
Zadaje pytania notatnikom Google NotebookLM przez przeglądarkę i zwraca odpowiedzi z cytatami z własnych źródeł, bez wyszukiwania w sieci.
- Autor
- akillness ↗
- Repozytorium
- jeo-skills ↗
- Licencja
- MIT
- Ostatnia zmiana repozytorium
- 13.11.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 5
- Skrypty
- 1
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 4
Brak sygnałów. Repozytorium zmienione 13.11.2026, bez odwołań do wycofanych modeli i API.
MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.
- 1 skrypt w paczce. Przejrzyj je przed instalacją, bo Claude może je uruchomić.
Skrypty: scripts/run.py
Co robi i kiedy
Zadaje pytania notatnikom Google NotebookLM przez przeglądarkę i zwraca odpowiedzi z cytatami z własnych źródeł, bez wyszukiwania w sieci.
Oryginalny opis (EN)
Queries managed Google NotebookLM notebooks for citation-backed, source-grounded answers via local Claude Code browser automation; use for uploaded sources, not live web search.
Jak działa
- Installation
- Requirements
- Setup (first-time)
- When to use this skill
- Do not use when
- Usage
- Data storage
- Key constraints
- Troubleshooting
- Follow-up protocol
Podgląd SKILL.md
--- name: notebooklm description: >- Queries managed Google NotebookLM notebooks for citation-backed, source-grounded answers via local Claude Code browser automation; use for uploaded sources, not live web search. allowed-tools: Bash Read Write Glob Grep compatibility: > Local Claude Code only — web UI is not supported (sandbox restrictions prevent network access). Requires Python 3.10+, Chrome (installed via `patchright install chrome`), and a Google account. Free Google accounts are limited to 50 queries/day. license: MIT metadata: tags: notebooklm, google, browser-automation, patchright, research, notebook, knowledge-base, citation, grounded-answers platforms: Claude Code keyword: notebooklm version: "1.3.0" source: PleasePrompto/notebooklm-skill modernization: 2026-05-04 --- # notebooklm — Google NotebookLM Integration for Claude Code > Query your NotebookLM notebooks directly. Source-grounded, citation-backed answers without leaving the editor. `notebooklm` brings Google NotebookLM into your Claude Code workflow via browser automation (Patchright). Instead of switching tabs and copy-pasting, you ask questions from the terminal and get grounded answers from your uploaded sources. ## Installation ### Plugin (Claude Code — recommended) ```bash claude plugin marketplace add PleasePrompto/notebooklm-skill ``` ### Manual (git clone) ```bash mkdir -p ~/.claude/skills cd ~/.claude/skills git clone https://github.com/PleasePrompto/notebooklm-skill.git notebooklm ``` ### Skill (via skills CLI) ```bash npx skills add https://github.com/akillness/jeo-skills --skill notebooklm ``` ## Requirements - **Python 3.10+** — for browser automation scripts - **Chrome** — installed via `patchright install chrome` (Chromium will NOT work) - **Google account** — a dedicated account is recommended for automation - **Local Claude Code** — web UI is not supported (sandbox restrictions) - **Internet access** — required to reach notebooklm.google.com ## Setup (first-time) ```bash # 1. Install Python dependencies (auto-runs on first use) cd ~/.claude/skills/notebooklm python scripts/run.py auth_manager.py setup # 2. A Chrome window opens — log in to Google manually # The session is saved and reused for future queries # 3. Add your first notebook python scripts/run.py notebook_manager.py add \ --url "https://notebooklm.google.com/notebook/YOUR_ID" \ --name "my-research" \ --description "Research notes on topic X" ``` ## When to use this skill - You have documents uploaded to NotebookLM and need citation-backed answers during development - You want to cross-reference your research notes while writing code without switching windows - You need source-grounded analysis from a curated document set (not a general web search) - You are doing research synthesis across multiple notebooks on related topics ## Do not use when - You need live web search — use `scrapling` or `web-research` instead - You need general knowledge not in your uploaded documents — ask Claude directly - You are running Claude from the web UI — browser automation is blocked in the sandbox - Your Google account has exceeded the 50 queries/day free-tier limit ## Usage ### Ask a question ```bash # Ask the currently active notebook python scripts/run.py ask_question.py --question "What are the key findings about X?" # Ask a specific notebook by ID python scripts/run.py ask_question.py \ --notebook-id "my-research" \ --question "Summarize the methodology section" # Ask with browser visible (useful for debugging) python scripts/run.py ask_question.py \ --question "What does source 3 say about Y?" \ --show-browser ``` ### Manage notebooks ```bash # List all notebooks python scripts/run.py notebook_manager.py list # Search notebooks python scripts/run.py notebook_manager.py search "machine learning" # Activate a notebook (sets it as the default) python scripts/run.py notebook_manager.py activate --id "my-research" # Remove a notebook from the library python scripts/run.py notebook_manager.py remove --id "old-notes" ``` ### Authentication management ```bash # Check auth status python scripts/run.py auth_manager.py status # Re-authenticate if session expired python scripts/run.py auth_manager.py reauth # Clear all auth data python scripts/run.py auth_manager.py clear ``` ## Data storage All user data is stored locally at `~/.claude/skills/notebooklm/data/`: ``` data/ ├── library.json # Your notebook registry ├── auth_info.json # Authentication metadata ├── browser_state/ # Persistent browser session │ ├── browser_profile/ # Chrome profile data │ └── state.json # Cookies and localStorage └── sessions.json # Active session tracking ``` ## Key constraints | Constraint | Detail | |------------|--------| | **Rate limit** | 50 queries/day (free Google account) | | **Local only** | Web UI sandbox blocks network access | | **Manual upload** | Documents must be uploaded to NotebookLM manually first | | **Browser overhead** | Each query opens a browser session (adds 3–10 seconds) | | **Auth renewal** | Session cookies expire; run `auth_manager.py reauth` when needed | ## Troubleshooting | Symptom | Fix | |---------|-----| | `Authentication failed` | Run `python scripts/run.py auth_manager.py reauth` | | `Browser crash` | Run `python scripts/run.py auth_manager.py clear` then `setup` | | `No response / timeout` | Increase timeout or run with `--show-browser` to inspect | | `Rate limited` | Wait 24 hours or use a different Google account | | `Chrome not found` | Run `patchright install chrome` (not `chromium`) | | `Import error` | First-run auto-setup may have failed; run `setup_environment.py` manually | Full troubleshooting guide: `~/.claude/skills/notebooklm/references/troubleshooting.md` ## Follow-up protocol After each answer, always assess completeness: > **"Is that ALL you need to know from this notebook on this topic?"** NotebookLM can answer follow-up questions within the same session context. Ask targeted follow-ups before switching notebooks to get the most grounded coverage. ## References - [PleasePrompto/notebooklm-skill](https://github.com/PleasePrompto/notebooklm-skill) — source repository - `~/.claude/skills/notebooklm/references/usage_patterns.md` — advanced usage patterns - `~/.claude/skills/notebooklm/references/api_reference.md` — programmatic usage - `~/.claude/skills/notebooklm/references/troubleshooting.md` — troubleshooting guide - `~/.claude/skills/notebooklm/AUTHENTICATION.md` — hybrid auth system details ## Instructions 1. Identify the task trigger and expected output. 2. Follow the workflow steps in this skill from top to bottom. 3. Validate outputs before moving to the next step. 4. Capture blockers and fallback path if any step fails. ## Examples - Example: Apply this skill to a small scope first, then scale to full scope after validation passes. ## Best practices - Keep outputs deterministic and auditable. - Prefer small reversible changes over broad risky edits. - Record assumptions explicitly.
Pliki w paczce (5)
- SKILL.md 7 KB
- references/api_reference.md 1 KB
- references/troubleshooting.md 1 KB
- references/usage_patterns.md 1 KB
- scripts/run.py 1 KB
Wersje
| Repozytorium | Wynik | Licencja |
|---|---|---|
| Ta strona · akillness/jeo-skills | 100 | MIT |
| rootcastleco/rei-skills | 100 | MIT |
| dreamlab-ai/agentbox | 100 | AGPL-3.0 |
| alirezarezvani/claude-skills | 100 | MIT |
| mathews-tom/armory | 100 | MIT |
| roomi-fields/notebooklm-mcp | 100 | MIT |
| llm011/ethan-agent | 100 | MIT |
| ArtemXTech/personal-os-skills | 100 | MIT |
| julianobarbosa/claude-code-skills | 100 | MIT |
Migawka bazy z 13.11.2026 · wersja paczki ea9c2c997fb8 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.
Podobne skille
- 95nblm
Odpytuje notatniki Google NotebookLM z Claude Code i zwraca odpowiedzi Gemini z cytowaniami, oparte wyłącznie na dokumentach, z automatyzacją przeglądarki, zarządzaniem biblioteką i trwałym uwierzytelnianiem. - 100doc-notebooklm-control
Automatyzuje w przeglądarce prace w NotebookLM, w tym notatniki, źródła i wyniki Studio. - 100blog-notebooklm
Odpytuje notatniki Google NotebookLM o odpowiedzi z cytatami z wgranych dokumentów, zarządza biblioteką notatników i uwierzytelnianiem Google oraz wspiera inteligentne wyszukiwanie. - 100blog-notebooklm
Odpytuje notatniki Google NotebookLM o odpowiedzi z cytatami z wgranych dokumentów, zarządza biblioteką notatników i uwierzytelnianiem Google oraz dostarcza kontekst researchu opartego na źródłach. - 100notebooklm-cli
Steruje Google NotebookLM przez CLI cli-web-notebooklm: tworzenie notatników, dodawanie źródeł URL i tekstu, pytania oparte na źródłach oraz generowanie i pobieranie podcastów, wideo, slajdów, quizów, FAQ i infografik. - 100notebooklm-connector
Łączy Claude Code i Codex CLI z własnym Google NotebookLM przez MCP do pytań o treści, z opcjonalnym ExportKit do eksportu lokalnego i nlm CLI do zarządzania źródłami.