NotebookLM стал в 10 раз лучше (AntiGravity)

# Шаг 1: Установка пакета Установите `notebooklm-mcp-server` через pip: ```bash pip install notebooklm-mcp-server ``` ## Шаг 2: Добавление в конфигурацию MCP Откройте файл конфигурации MCP (обычно `~/.gemini/antigravity/mcp_config.json` или `~/.config/opencode/opencode.json`) и добавьте следующую секцию в объект `mcpServers`: ```json "notebooklm": { "command": "notebooklm-mcp", "args": [], "env": {}, "disabledTools": [] } ``` **Полный пример конфигурации:** ```json { "mcpServers": { "notebooklm": { "command": "notebooklm-mcp", "args": [], "env": {}, "disabledTools": [] } } } ``` > **ВАЖНО:** Используйте команду `notebooklm-mcp`, а НЕ `python -m notebooklm_mcp_server`! ## Шаг 3: Авторизация в NotebookLM Запустите команду авторизации: ```bash notebooklm-mcp-auth ``` Эта команда: 1. Откроет Chrome с профилем для NotebookLM 2. Попросит вас войти в Google аккаунт (если ещё не вошли) 3. Автоматически извлечёт cookies и CSRF токен 4. Сохранит токены в `~/.notebooklm-mcp/auth.json` **Ожидаемый результат:** ``` SUCCESS! ======================================== Cookies: 29 extracted CSRF Token: Yes Session ID: boq_labs-tailwind-frontend_... ``` ## Шаг 4: Перезапуск AI ассистента Перезапустите Antigravity (или другой AI ассистент), чтобы MCP сервер загрузился. ## Шаг 5: Проверка работы После перезапуска попросите AI: ``` Покажи мои ноутбуки NotebookLM ``` или ``` List my NotebookLM notebooks ``` Если всё работает, вы увидите список ваших ноутбуков. ## Возможные проблемы ### Проблема: "command not found: notebooklm-mcp" **Решение:** Убедитесь, что Python Scripts находится в PATH: - Windows: `C:\Users\<USERNAME>\AppData\Local\Programs\Python\Python3XX\Scripts` - Linux/Mac: `~/.local/bin` ### Проблема: "server name notebooklm not found" **Решение:** Перезапустите AI ассистент после добавления конфигурации. ### Проблема: Ошибки авторизации **Решение:** Повторите команду `notebooklm-mcp-auth` и убедитесь, что вы вошли в Google аккаунт в открывшемся Chrome. ## Доступные возможности После установки вы можете: - ✅ Просматривать список ноутбуков - ✅ Получать информацию о ноутбуках и источниках - ✅ Добавлять источники (URL, Google Drive, текст) - ✅ Задавать вопросы по содержимому ноутбуков - ✅ Создавать отчёты, аудио-обзоры, инфографику - ✅ Генерировать флешкарты, квизы, таблицы данных - ✅ Проводить исследования (web/Drive search) ## Полезные команды ```bash # Проверка установки pip show notebooklm-mcp-server # Проверка доступности команды notebooklm-mcp --help # Повторная авторизация notebooklm-mcp-auth # Проверка версии pip list | grep notebooklm ``` --- **Версия инструкции:** 1.0 **Дата:** 01.02.2026 **Протестировано на:** Windows 11, Python 3.13

ШАБЛОН НАВЫКА #Antigravity Skill Creator System Instructions You are an expert developer specializing in creating "Skills" for the Antigravity agent environment. Your goal is to generate high-quality, predictable, and efficient .agent/skills/ directories based on user requirements. 1. Core Structural Requirements Every skill you generate must follow this folder hierarchy: * / * SKILL.md (Required: Main logic and instructions) * scripts/ (Optional: Helper scripts) * examples/ (Optional: Reference implementations) * resources/ (Optional: Templates or assets) 2. YAML Frontmatter Standards The SKILL.md must start with YAML frontmatter following these strict rules: * name: Gerund form (e.g., testing-code, managing-databases). Max 64 chars. Lowercase, numbers, and hyphens only. No "claude" or "anthropic" in the name. * description: Written in third person. Must include specific triggers/keywords. Max 1024 chars. (e.g., "Extracts text from PDFs. Use when the user mentions document processing or PDF files.") 3. Writing Principles (The "Claude Way") When writing the body of SKILL.md, adhere to these best practices: * Conciseness: Assume the agent is smart. Do not explain what a PDF or a Git repo is. Focus only on the unique logic of the skill. * Progressive Disclosure: Keep SKILL.md under 500 lines. If more detail is needed, link to secondary files (e.g., [See ADVANCED.md](ADVANCED.md)) only one level deep. * Forward Slashes: Always use / for paths, never \\. * Degrees of Freedom: * Use Bullet Points for high-freedom tasks (heuristics). * Use Code Blocks for medium-freedom (templates). * Use Specific Bash Commands for low-freedom (fragile operations). 4. Workflow & Feedback Loops For complex tasks, include: 1. Checklists: A markdown checklist the agent can copy and update to track state. 2. Validation Loops: A "Plan-Validate-Execute" pattern. (e.g., Run a script to check a config file BEFORE applying changes). 3. Error Handling: Instructions for scripts should be "black boxes"—tell the agent to run --help if they are unsure. 5. Output Template When asked to create a skill, output the result in this format: [Folder Name] * Path:* .agent/skills/[skill-name]/ [SKILL.md] --- name: [gerund-name] description: [3rd-person description] --- # [Skill Title] ## When to use this skill - [Trigger 1] - [Trigger 2] ## Workflow [Insert checklist or step-by-step guide here] ## Instructions [Specific logic, code snippets, or rules] ## Resources - [Link to scripts/ or resources/] **[Supporting Files]** (If applicable, provide the content for scripts/ or examples/) --- ## Instructions for use 1. **Copy the content above** into a new file named `antigravity-skill-creator.md`. 2. **Upload this file** to your AI agent or paste it into the system prompt area. 3. **Trigger a skill creation** by saying: *"Based on my skill creator instructions, build me a skill for [Task, e.g., 'automating React component testing with Vitest']."* ### Suggested Next Step Would you like me to use this new logic to **generate a specific example skill** for you right now (such as a "Deployment Guard" or "Code Reviewer" skill)? ВОЗМОЖНОСТИ НОТЕБУКЛМ MCP: notebooklm Configure Enabled 1. refresh_auth Reload auth tokens from disk or run headless re-authentication. Call this after running notebooklm-mcp-auth to pick up new tokens, or to attempt automatic re-authentication if Chrome profile has saved login. Returns status indicating if tokens were refreshed successfully. 2. notebook_list List all notebooks. Args: max_results: Maximum number of notebooks to return (default: 100) 3. notebook_create Create a new notebook. Args: title: Optional title for the notebook 4. notebook_get Get notebook details with sources. Args: notebook_id: Notebook UUID 5. notebook_describe Get AI-generated notebook summary with suggested topics. Args: notebook_id: Notebook UUID Returns: summary (markdown), suggested_topics list 6. source_describe Get AI-generated source summary with keyword chips. Args: source_id: Source UUID Returns: summary (markdown with **bold** keywords), keywords list 7. source_get_content Get raw text content of a source (no AI processing). Returns the original indexed text from PDFs, web pages, pasted text, or YouTube transcripts. Much faster than notebook_query for content export. Args: source_id: Source UUID Returns: content (str), title (str), source_type (str), char_count (int) 8. notebook_add_url Add URL (website or YouTube) as source. Args: notebook_id: Notebook UUID url: URL to add 9. notebook_add_text Add pasted text as source. Args: notebook_id: Notebook UUID text: Text content to add title: Optional title 10. notebook_add_drive Add Google Drive document as source. Args: notebook_id: Notebook UUID document_id: Drive document ID (from URL) title: Display title doc_type: doc|slides|sheets|pdf 11. notebook_query Ask AI about EXISTING sources already in notebook. NOT for finding new sources. Use research_start instead for: deep research, web search, find new sources, Drive search. Args: notebook_id: Notebook UUID query: Question to ask source_ids: Source IDs to query (default: all) conversation_id: For follow-up questions timeout: Request timeout in seconds (default: from env NOTEBOOKLM_QUERY_TIMEOUT or 120.0) 12. notebook_delete Delete notebook permanently. IRREVERSIBLE. Requires confirm=True. Args: notebook_id: Notebook UUID confirm: Must be True after user approval 13. notebook_rename Rename a notebook. Args: notebook_id: Notebook UUID new_title: New title 14. chat_configure Configure notebook chat settings. Args: notebook_id: Notebook UUID goal: default|learning_guide|custom custom_prompt: Required when goal=custom (max 10000 chars) response_length: default|longer|shorter 15. source_list_drive List sources with types and Drive freshness status. Use before source_sync_drive to identify stale sources. Args: notebook_id: Notebook UUID 16. source_sync_drive Sync Drive sources with latest content. Requires confirm=True. Call source_list_drive first to identify stale sources. Args: source_ids: Source UUIDs to sync confirm: Must be True after user approval 17. source_delete Delete source permanently. IRREVERSIBLE. Requires confirm=True. Args: source_id: Source UUID to delete confirm: Must be True after user approval 18. research_start Deep research / fast research: Search web or Google Drive to FIND NEW sources. Use this for: "deep research on X", "find sources about Y", "search web for Z", "search Drive". Workflow: research_start -> poll research_status -> research_import. Args: query: What to search for (e.g. "quantum computing advances") source: web|drive (where to search) mode: fast (~30s, ~10 sources) | deep (~5min, ~40 sources, web only) notebook_id: Existing notebook (creates new if not provided) title: Title for new notebook 19. research_status Poll research progress. Blocks until complete or timeout. Args: notebook_id: Notebook UUID poll_interval: Seconds between polls (default: 30) max_wait: Max seconds to wait (default: 300, 0=single poll) compact: If True (default), truncate report and limit sources shown to save tokens. Use compact=False to get full details. task_id: Optional Task ID to poll for a specific research task. 20. research_import Import discovered sources into notebook. Call after research_status shows status="completed". Args: notebook_id: Notebook UUID task_id: Research task ID source_indices: Source indices to import (default: all) 21. audio_overview_create Generate audio overview. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) format: deep_dive|brief|critique|debate length: short|default|long language: BCP-47 code (en, es, fr, de, ja) focus_prompt: Optional focus text confirm: Must be True after user approval 22. video_overview_create Generate video overview. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) format: explainer|brief visual_style: auto_select|classic|whiteboard|kawaii|anime|watercolor|retro_print|heritage|paper_craft language: BCP-47 code (en, es, fr, de, ja) focus_prompt: Optional focus text confirm: Must be True after user approval 23. studio_status Check studio content generation status and get URLs. Args: notebook_id: Notebook UUID 24. studio_delete Delete studio artifact. IRREVERSIBLE. Requires confirm=True. Args: notebook_id: Notebook UUID artifact_id: Artifact UUID (from studio_status) confirm: Must be True after user approval 25. infographic_create Generate infographic. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) orientation: landscape|portrait|square detail_level: concise|standard|detailed language: BCP-47 code (en, es, fr, de, ja) focus_prompt: Optional focus text confirm: Must be True after user approval 26. slide_deck_create Generate slide deck. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) format: detailed_deck|presenter_slides length: short|default language: BCP-47 code (en, es, fr, de, ja) focus_prompt: Optional focus text confirm: Must be True after user approval 27. report_create Generate report. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) report_format: "Briefing Doc"|"Study Guide"|"Blog Post"|"Create Your Own" custom_prompt: Required for "Create Your Own" language: BCP-47 code (en, es, fr, de, ja) confirm: Must be True after user approval 28. flashcards_create Generate flashcards. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) difficulty: easy|medium|hard confirm: Must be True after user approval 29. quiz_create Generate quiz. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) question_count: Number of questions (default: 2) difficulty: Difficulty level (default: medium) confirm: Must be True after user approval 30. data_table_create Generate data table. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID description: Description of the data table to create source_ids: Source IDs (default: all) language: Language code (default: "en") confirm: Must be True after user approval 31. mind_map_create Generate and save mind map. Requires confirm=True after user approval. Args: notebook_id: Notebook UUID source_ids: Source IDs (default: all) title: Display title confirm: Must be True after user approval 32. save_auth_tokens Save NotebookLM cookies (FALLBACK method - try notebooklm-mcp-auth first!). IMPORTANT FOR AI ASSISTANTS: - First, run `notebooklm-mcp-auth` via Bash/terminal (automated, preferred) - Only use this tool if the automated CLI fails Args: cookies: Cookie header from Chrome DevTools (only needed if CLI fails) csrf_token: Deprecated - auto-extracted session_id: Deprecated - auto-extracted request_body: Optional - contains CSRF if extracting manually request_url: Optional - contains session ID if extracting manually