NotebookLM CLI

ReviewAudited by ClawScan on May 10, 2026.

Overview

This is a transparent NotebookLM CLI wrapper, but installing it lets the agent use an authenticated NotebookLM account to change, delete, export, or share notebook content when invoked.

This skill appears coherent and purpose-aligned, but treat it like giving the agent access to your NotebookLM CLI. Verify the external `notebooklm` binary, log in with the correct account, and require confirmation before deleting content, exporting artifacts, changing collaborators, or enabling public sharing.

Findings (5)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

If invoked incorrectly, the agent could delete notebook data or make notebook content publicly accessible.

Why it was flagged

The command catalog exposes destructive and public-sharing NotebookLM operations through the wrapper. These are purpose-aligned for a CLI wrapper, but they are high-impact actions that should be user-approved before execution.

Skill content
node {baseDir}/scripts/notebooklm.mjs delete --notebook <notebook_id> --yes
node {baseDir}/scripts/notebooklm.mjs source delete <source_id> --yes
node {baseDir}/scripts/notebooklm.mjs share public --enable
Recommendation

Use explicit user confirmation for delete, export, permission-change, and public-sharing commands.

What this means

The skill will run whatever `notebooklm` executable the environment resolves first.

Why it was flagged

The skill executes the external `notebooklm` CLI. This is expected for the stated wrapper purpose, but users should ensure the intended binary is installed on PATH.

Skill content
const result = spawnSync("notebooklm", args, { stdio: "inherit" });
Recommendation

Install the NotebookLM CLI from a trusted source and verify PATH order before use.

What this means

Agent actions may read or modify NotebookLM data under the logged-in account.

Why it was flagged

The skill expects the underlying CLI to be authenticated, so commands run with the user’s NotebookLM account authority. This is expected for the integration and no credential leakage is shown.

Skill content
NotebookLM CLI authenticated (run `login` if needed).
Recommendation

Authenticate only the intended account and review sensitive account-changing commands before allowing execution.

What this means

Users must independently verify the external CLI they install and run.

Why it was flagged

The metadata does not declare the external `notebooklm` dependency or provide its source, even though SKILL.md requires it on PATH. This is an incomplete provenance signal rather than evidence of malicious behavior.

Skill content
Source: unknown; Homepage: none; Install specifications: No install spec; Required binaries (all must exist): none
Recommendation

Document the trusted NotebookLM CLI source and declare required binaries in metadata.

What this means

Sensitive or incorrect content added to a notebook may persist and be reused in future NotebookLM interactions.

Why it was flagged

The skill can add and update persistent NotebookLM sources/notes and retrieve notebook history. This is normal NotebookLM behavior, but stored context may influence later answers.

Skill content
node {baseDir}/scripts/notebooklm.mjs source add "Inline notes" --title "Meeting" --notebook <notebook_id>
node {baseDir}/scripts/notebooklm.mjs note save <note_id> --content "Updated notes"
node {baseDir}/scripts/notebooklm.mjs history --notebook <notebook_id> --limit 20
Recommendation

Add only intended content to notebooks and periodically review sources, notes, and sharing settings.