notebooklm

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

• Third-party content exposure detected (medium risk: 0.30). In workflows/youtube-channel.md → scripts/load_channel.py load the workflow ingests YouTube video transcripts fetched from the user-supplied --channel URL and later uses them in nlm notebook query (workflows/ask.md) to generate free-text answers.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's installer runs a remote installer via curl | sh (https://astral.sh/uv/install.sh) inside scripts/bootstrap.sh and the README explicitly instructs cloning the skill repo (git clone https://github.com/artemnovitckii/notebooklm-coach.git) as part of install — these fetch-and-execute steps download and run external code during setup.