Notebooklm
WarnAudited by ClawScan on May 18, 2026.
Overview
The skill matches a NotebookLM integration, but it relies on an unofficial unreviewed CLI tied to your Google account and allows some account/file-import actions without confirmation.
Before installing, verify the exact `notebooklm-py` version and installation path, use a dedicated or low-risk Google account if possible, and require confirmation before uploading sources, switching profiles, changing settings, sharing, or deleting content.
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.
You could be asked to authenticate your Google account through an external CLI/library whose exact installed version and contents are not visible in this skill package.
The provided package has no install spec or code files, so the executable/library that would handle Google auth and API calls is not pinned or reviewable in the supplied artifacts.
> OpenClaw handles installation and setup automatically. ... Source library: [teng-lin/notebooklm-py] ... **⚠️ Unofficial** — Uses undocumented Google APIs.
Require a pinned, reviewable install spec or manually verify the upstream library and version before logging in; consider using a dedicated Google account.
The CLI may be able to act as you in NotebookLM and related Google flows used by NotebookLM, including notebook/source operations.
The skill requires delegated Google account access, while the registry metadata declares no primary credential and the artifacts do not specify scopes, session storage, or account-access limits.
Authentication must be active before any command: `notebooklm login` # Opens browser — log in with Google account
Only authenticate if you trust the exact CLI being installed; prefer a least-privilege or test Google account and review Google account/session permissions.
After activation, the agent may create notebooks, add sources, or switch profiles/settings without pausing for explicit approval for each change.
The no-confirmation list includes actions that mutate notebook/account/tool state, not only read-only checks. Source addition can import user content into NotebookLM, and profile/language operations can change execution context.
**Run without confirmation:** ... `notebooklm create` ... `notebooklm use <id>` ... `notebooklm source add` ... `notebooklm language list/get/set` ... `notebooklm profile list/create/switch`
Require confirmation for source imports, profile changes, language changes, and other account-state mutations, especially when files or private research are involved.
Private documents or chat outputs added to NotebookLM may remain in notebooks and be reused in later answers or shared workflows.
The skill persists sources, notes, and generated context in NotebookLM and can retrieve indexed source text; this is expected for the purpose but affects data retention and reuse.
| 📂 Add Sources | URLs, PDFs, YouTube, Google Drive, audio/video/images | ... | 📝 Notes | Save answers and history as notebook notes | ... ✅ Source full-text access
Avoid adding highly sensitive files unless you understand NotebookLM retention and sharing settings; periodically review and delete sources or notes you no longer need.
In multi-agent or parallel sessions, sources, questions, or generated content could be routed to an unintended notebook.
The skill itself notes that a shared active-notebook context can be unsafe in parallel use, because later commands may apply to the wrong notebook unless explicitly scoped.
`notebooklm use <id>` — set context (⚠️ single-agent only; use `-n` in parallel)
Use explicit notebook IDs or the `-n` option for parallel workflows, and confirm the active notebook before mutating actions.