notebooklm

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow can ingest outsider-authored free text when the user adds public web URLs or performs source add-research (web/drive discovery), and that fetched page/article text is then used as LLM context for ask/generation.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The setup instructions include commands that fetch and execute remote code at install time—specifically the pip install from a GitHub source (pip install "git+https://github.com/teng-lin/notebooklm-py@${LATEST_TAG}" together with the curl https://api.github.com/... used to obtain the tag) and the NPX installer (npx skills add teng-lin/notebooklm-py)—which will download and run third-party code as a required installation step.