連接你的notebookLM_AI效率大爆發

你會學到

• 透過「懶人包」與 Claude Code 桌面版，可以簡化 Google NotebookLM 的串接與操作流程。
• 即使不熟悉底層工具，也能利用 Claude Code 桌面版自動化執行 NotebookLM 的相關任務。
• Claude Code 透過模擬瀏覽器操作而非 API 方式，實現與 NotebookLM 的互動，並能自動產生多種格式的教學素材。
• 首次設定完成後，日後操作不需重複設定，大幅提升工作效率。
• 整合後可直接在 Claude Code 中下指令，讓 NotebookLM 自動將產出內容分類儲存至指定資料夾。

跟著做

1. 從影片提供的連結下載 NotebookLM 懶人包（Markdown 格式檔案）。
2. 在電腦中（建議 C 槽使用者資料夾下）建立一個專屬的本地資料夾，例如 notebook LM。
3. 開啟 Claude Code 桌面版，新建一個項目並將資料夾路徑指向步驟 2 建立的資料夾。
4. 將下載的懶人包 Markdown 檔案拖曳至 Claude Code 中並提交。
5. 同意信任此 Workspace。
6. 在 Claude Code 設定中，開啟「Bypass Permissions」選項。
7. 等待 Claude Code 自動檢查並安裝所需工具（如 Git for Windows, uv, mcp cli）。
8. 依提示在瀏覽器中登入 Google 帳號進行驗證。
9. 完成登入後，關閉並重新啟動 Claude Code 桌面版。
10. 返回原有對話，告知 Claude Code 已重啟，讓它繼續執行。
11. Claude Code 將自動建立簡報、資訊圖表、Google 文件等多個預設輸出資料夾。
12. 測試功能，直接在 Claude Code 中輸入指令，如「將今天安裝懶人包的整個過程丟到 notebook LM做成資訊圖表」。

教學金句

「從這一集開始，徹底解放你的 Claude Code 能力。」 「你不需要懂這些工具的用途是什麼。」 「所有的操作只要設定過一次之後就不用再設定。」 「目前 NotebookLM 是沒有 API 的，所以我們沒辦法直接呼叫 NLM，是用模擬器瀏覽操作方式。」 「如果你跟著做到這邊的話，恭喜你，你已經完成了 Claude 串接 NotebookLM。」

適用情境

這部影片適合具備 Claude Code 桌面版基礎使用經驗的學習者，可作為 AI 教學課程中「AI 工具整合與自動化」或「大型語言模型與知識庫連結應用」單元的一部分。它示範了如何將本地 AI 開發環境與雲端筆記本工具連結，實現內容自動化生成與管理。學員需具備 Claude Code EP01、EP02 的前置知識，以及基本的軟體安裝與檔案管理能力。

延伸主題

• 深入探索 MCP（Multi-modal Control Protocol）的架構與自訂擴展方法。
• 利用 Claude Code 與 NotebookLM 進行垂直領域知識萃取與內容自動化生產實戰。
• NotebookLM 筆記本管理、資料來源整合及協作功能的進階應用。
• 比較其他主流 AI 助手與 NotebookLM 整合的技術路徑與優劣。

---en---

What You’ll Learn

• Through the “lazy pack” and Claude Code desktop version, the connection and operation process of Google NotebookLM can be simplified.
• Even without familiarity with underlying tools, Claude Code desktop can automate NotebookLM related tasks.
• Claude Code interacts with NotebookLM by simulating browser operations rather than API calls, and can automatically generate teaching materials in various formats.
• Once the initial setup is complete, future operations do not require repetitive configuration, significantly improving work efficiency.
• After integration, commands can be directly issued in Claude Code, allowing NotebookLM to automatically classify and save generated content into specified folders.

Follow These Steps

1. Download the NotebookLM lazy pack (Markdown format file) from the link provided in the video.
2. Create a dedicated local folder on your computer (preferably under the C drive user folder), for example, notebook LM.
3. Open Claude Code desktop, create a new project, and point the folder path to the folder created in step 2.
4. Drag and drop the downloaded lazy pack Markdown file into Claude Code and submit it.
5. Agree to trust this Workspace.
6. In Claude Code settings, enable the “Bypass Permissions” option.
7. Wait for Claude Code to automatically check and install necessary tools (such as Git for Windows, uv, mcp cli).
8. Follow the prompts to log in to your Google account in the browser for verification.
9. After successful login, close and restart Claude Code desktop.
10. Return to the original conversation, inform Claude Code that it has restarted, and let it continue execution.
11. Claude Code will automatically create multiple default output folders for presentations, infographics, and Google Docs.
12. Test the functionality by directly inputting commands into Claude Code, such as “throw the entire process of installing the lazy pack today into notebook LM and make it into an infographic.”

Teaching Quote

「From this episode, completely unleash the power of your Claude Code.」 「You don’t need to understand what these tools are for.」 「All operations only need to be set up once and never again.」 「Currently, NotebookLM does not have an API, so we cannot directly call NLM; we use a browser simulation method.」 「If you have followed along this far, congratulations, you have completed the Claude-NotebookLM connection.」

When to Use

This video is suitable for learners with basic experience in using Claude Code desktop version, and can be used as part of an AI teaching course in the “AI Tool Integration and Automation” or “Large Language Model and Knowledge Base Link Application” units. It demonstrates how to link a local AI development environment with a cloud notebook tool to achieve automated content generation and management. Learners should have prior knowledge of Claude Code EP01, EP02, and basic software installation and file management skills.

Related Topics

• In-depth exploration of the architecture and custom extension methods of MCP (Multi-modal Control Protocol).
• Practical application of Claude Code and NotebookLM for domain-specific knowledge extraction and automated content production.
• Advanced applications of NotebookLM for notebook management, data source integration, and collaboration features.
• Comparison of technical paths and pros/cons of integrating other mainstream AI assistants with NotebookLM.