← 返回 Skills 市场
NotebookLM Studio
作者
jasontsaicc
· GitHub ↗
· v2.1.3
· MIT-0
268
总下载
1
收藏
1
当前安装
3
版本数
在 OpenClaw 中安装
/install notebooklm-studio
功能描述
Import sources (URLs, YouTube, files, text) into Google NotebookLM and generate user-selected artifacts: podcast, video, report, quiz, flashcards, mind map,...
安全使用建议
This skill appears coherent for driving the unofficial notebooklm CLI and delivering artifacts (audio/video require ffmpeg). Before installing: 1) Verify you trust the notebooklm-py package source and review its code (it controls authentication and generation). 2) Be careful when following the headless-server instructions that ask you to copy ~/.notebooklm/storage_state.json — this file contains session credentials; transfer it only over secure channels, set strict file permissions, and prefer ephemeral or single-tenant servers. 3) If you enable the suggested cron recovery or health-checks, run them in an isolated account or container to limit exposure. 4) Confirm Telegram delivery behavior and that your agent/platform message tool will not leak artifacts to unintended recipients. 5) If you need stronger guarantees, run the skill only on a trusted machine, or avoid copying session credentials to shared/production hosts.
功能分析
Type: OpenClaw Skill
Name: notebooklm-studio
Version: 2.1.3
The notebooklm-studio skill bundle is a well-structured automation tool for Google NotebookLM. It implements a sophisticated 9-step workflow in SKILL.md, including a two-tier generation strategy to handle long-running tasks like audio and video generation. The bundle includes functional shell scripts (scripts/compress_audio.sh and scripts/recover_tier2_delivery.sh) for post-processing and task recovery via cron, which are transparently coded and aligned with the stated purpose. No evidence of data exfiltration, malicious command execution, or prompt injection was found; the logic is focused entirely on notebook management and artifact delivery.
能力评估
Purpose & Capability
Name/description (import sources → generate artifacts) matches the declared binaries (notebooklm, ffmpeg) and the SKILL.md workflow. The included scripts (audio compression, recovery) are appropriate for Telegram delivery and long-running artifact generation. Nothing required by the skill (binaries, files, or installation steps) is unrelated to the stated NotebookLM-focused purpose.
Instruction Scope
Instructions stay within the NotebookLM workflow (auth check, create/use notebook, add sources, generate artifacts, download, deliver). They explicitly require saving plain text to temp files, adding Google Drive file IDs, and using OpenClaw's message tool for Telegram delivery. Notable scope items to review: the README and SKILL.md recommend copying the NotebookLM browser storage_state.json between machines (moving session credentials), and they instruct creating/maintaining delivery-status.json and cron-driven recovery — all functionally justified but sensitive.
Install Mechanism
No opaque or remote install spec in the skill bundle; installation instructions point to pip (notebooklm-py), playwright, Homebrew/apt, and optional ClawHub/npm or git clone. These are standard package sources (pip, Playwright, GitHub). No downloads from shorteners/personal IPs or archive-extract steps in the provided files.
Credentials
The skill declares no required environment variables and only needs the notebooklm CLI and ffmpeg binaries — proportional to its tasks. Caveats: it instructs transferring the NotebookLM session file (~/.notebooklm/storage_state.json) for headless servers, which is effectively moving credentials and must be handled with caution (secure copy, strict file permissions). It also performs file I/O under ./output and can be run by cron, which gives persistent access to any artifacts the session can access.
Persistence & Privilege
The skill is not force-included (always: false) and does not request elevated platform privileges. However, the README and scripts recommend persistent cron jobs (recovery and auth health checks) and instruct symlinking into agent skill directories. These are operational recommendations (optional) that, if followed, grant ongoing background behavior tied to the notebooklm session. That persistence is plausible for the stated recovery use-case but is something to opt into consciously.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install notebooklm-studio - 安装完成后，直接呼叫该 Skill 的名称或使用
/notebooklm-studio触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v2.1.3
Add Chinese README, full flow screenshot, update setup instructions
v2.1.2
Add ClawHub install option, separate CLI dependency in README, add .clawignore
v2.1.1
Auth precheck, two-tier generation, delivery recovery, dedupgate
元数据
常见问题
NotebookLM Studio 是什么？
Import sources (URLs, YouTube, files, text) into Google NotebookLM and generate user-selected artifacts: podcast, video, report, quiz, flashcards, mind map,... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 268 次。
如何安装 NotebookLM Studio？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install notebooklm-studio」即可一键安装，无需额外配置。
NotebookLM Studio 是免费的吗？
是的，NotebookLM Studio 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
NotebookLM Studio 支持哪些平台？
NotebookLM Studio 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 NotebookLM Studio？
由 jasontsaicc（@jasontsaicc）开发并维护，当前版本 v2.1.3。
推荐 Skills