video-transcript

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

• Third-party content exposure detected (low risk: 0.10). The required workflow processes user-provided video URLs or local files by downloading them and sending their frames/audio to an AI model for transcription; since the agent processes selected or user-supplied URLs rather than continuously monitoring an attacker-controlled feed, the exposure risk is low.

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

• Potentially malicious external URL detected (medium risk: 0.30). The installation and bootstrap routines fetch script and packaging content from a raw GitHub user content URL and a GitHub repository archive, which represent unverified operational runtime and setup dependencies from an ordinary unofficial project repository.