← Back to Skills Marketplace
126
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install notebooklm-api
Description
Complete API for Google NotebookLM - full programmatic access including features not in the web UI. Create notebooks, add sources, generate all artifact type...
Usage Guidance
This skill appears to be a wrapper/guide for a NotebookLM CLI client and is not obviously malicious, but it has several mismatches and privacy risks you should consider before installing: 1) The SKILL.md references sensitive env vars (NOTEBOOKLM_AUTH_JSON) and a config directory (~/.notebooklm) but the registry metadata declares none—expect to provide and protect auth tokens (storage_state.json-like data). 2) The instructions allow the agent to run commands automatically (including `source add` and `create`) that can read local files and upload them; if you install, restrict automatic actions or require confirmations for any file uploads. 3) Use isolation: set NOTEBOOKLM_HOME to a dedicated directory (or /tmp/agent-$ID) when experimenting and do not expose auth JSON in plaintext in shared CI without secret management. 4) Verify the pip package source (use PyPI or a specific GitHub release tag as recommended) and inspect what `notebooklm skill install` does (it references a separate “Claude Code” skill). If you need stronger assurances, request the upstream package repository, the exact commands `notebooklm skill install` runs, or a declared list of env vars in the registry before proceeding.
Capability Analysis
Type: OpenClaw Skill
Name: notebooklm-api
Version: 0.3.2
The skill provides a comprehensive programmatic interface for Google NotebookLM via the `notebooklm-py` CLI tool. It includes detailed instructions for an AI agent to manage notebooks, sources, and artifact generation (e.g., podcasts, videos, and quizzes). While the skill handles sensitive Google authentication tokens and performs file/network operations, these actions are strictly aligned with its stated purpose. The documentation (SKILL.md) includes appropriate safety guidelines, such as requiring user confirmation for destructive actions (delete) or long-running tasks (download/generate), and provides clear diagnostic commands (`notebooklm auth check`) for troubleshooting. No evidence of malicious intent, data exfiltration, or harmful prompt injection was found.
Capability Assessment
Purpose & Capability
Name/description claim a NotebookLM API and the SKILL.md instructs installing a notebooklm CLI client and using Google OAuth—this is consistent. One oddity: SKILL.md tells the user to “install the Claude Code skill” after installing the client; that dependency is not explained in the registry metadata and is unexpected but may be an integration step.
Instruction Scope
The instructions reference reading/writing ~/.notebooklm (context.json), configuring NOTEBOOKLM_HOME, and using NOTEBOOKLM_AUTH_JSON (inline auth JSON). They list many commands that the agent may run automatically, including `notebooklm source add` (which can add/upload files or URLs) and `notebooklm create`. Allowing those to run without confirmation gives the agent ability to read local files and upload them to NotebookLM; the SKILL.md also advises storing auth tokens/ storage_state.json contents in an env var. These behaviors are within the claimed purpose but expand scope to accessing local files and persistent credentials.
Install Mechanism
This is an instruction-only skill; no install spec in the registry. The SKILL.md recommends installing from PyPI or a specific GitHub release tag and explicitly warns against using main—both reasonable and lower-risk than arbitrary downloads.
Credentials
Registry metadata shows no required env vars, but SKILL.md documents NOTEBOOKLM_HOME and NOTEBOOKLM_AUTH_JSON (sensitive) as important for CI and non-interactive auth. The skill should have declared these env vars and indicated that auth JSON contains tokens (storage_state.json). Asking for inline auth JSON without declaring it is a mismatch and a potential secret-handling risk.
Persistence & Privilege
The CLI writes persistent state to ~/.notebooklm and a context.json file; the SKILL.md warns about context collisions and recommends per-agent isolation. More concerning: the skill's autonomy rules permit many commands to run automatically (including adding sources), which may cause the agent to read/upload local files and persist auth context without explicit user confirmation. The skill is not marked always:true and doesn't change other skills, but its allowed automatic actions grant it broad file and network interactions.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install notebooklm-api - After installation, invoke the skill by name or use
/notebooklm-api - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.3.2
Full NotebookLM API access via notebooklm-py. Create notebooks, add sources, chat, generate audio/video/quiz/slides/infographics. Features beyond web UI: batch downloads, PPTX export, mind map JSON. No browser automation needed.
Metadata
Frequently Asked Questions
What is NotebookLM API?
Complete API for Google NotebookLM - full programmatic access including features not in the web UI. Create notebooks, add sources, generate all artifact type... It is an AI Agent Skill for Claude Code / OpenClaw, with 126 downloads so far.
How do I install NotebookLM API?
Run "/install notebooklm-api" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is NotebookLM API free?
Yes, NotebookLM API is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does NotebookLM API support?
NotebookLM API is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created NotebookLM API?
It is built and maintained by lmanchu (@lmanchu); the current version is v0.3.2.
More Skills