NotebookLM CLI

NotebookLM CLI wrapper via `node {baseDir}/scripts/notebooklm.mjs`. Use for auth, notebooks, chat, sources, notes, sharing, research, and artifact generation/download.

Security Scan

This is a thin, coherent wrapper that forwards arguments to the local 'notebooklm' CLI; it behaves as described and does not itself request extra privileges or install software.

Purpose & Capability

The skill claims to be a NotebookLM CLI wrapper and the included script simply spawns the 'notebooklm' binary with forwarded args. No unrelated binaries, env vars, or installs are requested — the requirements match the stated purpose.

Instruction Scope

SKILL.md instructs the agent to run the wrapper which in turn runs the system 'notebooklm' CLI with arbitrary arguments. The wrapper does not read files or env vars itself, but running the external CLI may perform any actions that CLI supports (including network access, reading local config, importing Drive files, etc.).

Install Mechanism

There is no install spec — this is instruction-only plus a tiny forwarding script. Nothing is written to disk by the skill beyond the provided script, so install risk is minimal.

Credentials

The skill declares no env vars, but it requires that the notebooklm CLI be installed and authenticated. That means the external CLI will use whatever credentials/configuration exist on the host (local token files, OAuth sessions, Drive access). The skill does not request those credentials explicitly, which is reasonable for a wrapper, but users should be aware the CLI will operate with their existing auth.

Persistence & Privilege

always is false and the skill contains no installation hooks or configuration changes. It does not modify other skills or system-wide settings. Note: model invocation is allowed (platform default), so an agent could autonomously run notebooklm commands if permitted by policy.

Assessment

This skill is a small, transparent wrapper that simply runs your installed 'notebooklm' CLI. Before installing: (1) confirm you trust the source of the notebooklm binary you have on PATH (the wrapper itself is tiny and readable), (2) be aware that any credentials/configs already set for the notebooklm CLI (including Drive or OAuth access) will be used by commands the agent runs, and (3) if you are concerned about an agent invoking commands autonomously, restrict autonomous invocation or review which commands the agent is allowed to run. If you want extra caution, inspect the notebooklm binary you will execute and ensure it is the official/expected implementation.

NotebookLM CLI Wrapper

Required parameters

• node and notebooklm available on PATH.
• NotebookLM CLI authenticated (run login if needed).

Quick start

• Wrapper script: scripts/notebooklm.mjs (invokes notebooklm CLI).
• Run from the skill directory or use an absolute {baseDir} path.

node {baseDir}/scripts/notebooklm.mjs status
node {baseDir}/scripts/notebooklm.mjs login
node {baseDir}/scripts/notebooklm.mjs list
node {baseDir}/scripts/notebooklm.mjs use <notebook_id>
node {baseDir}/scripts/notebooklm.mjs ask "Summarize the key takeaways" --notebook <notebook_id>

Request & output

• Command form: node {baseDir}/scripts/notebooklm.mjs <command> [args...].
• Prefer --json for machine-readable output.
• For long-running tasks, use --exec-timeout <seconds>; --timeout is reserved for wait/poll commands.

References

• references/cli-commands.md

Assets

• None.