notebooklm-research

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/setup_chrome.py uses subprocess.Popen to launch a local Chrome browser instance with specific debugging flags (--remote-debugging-port) and a dedicated user data directory. This is a functional requirement for the skill's stated purpose of automating NotebookLM without an API.
  • [DATA_EXPOSURE]: The skill explicitly advises the user to treat the ~/.notebooklm-chrome directory as a sensitive credential directory since it stores session tokens. The scripts themselves do not attempt to exfiltrate this data or access it outside of the Chrome browser process.
  • [REMOTE_CODE_EXECUTION]: While the skill installs and uses playwright, it does not download or execute unverified remote scripts. All automation logic is contained within the local notebooklm_client.py file.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 09:13 PM
Security Audit — agent-trust-hub — notebooklm-research