hooks

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

• Third-party content exposure detected (high risk: 0.95). Runtime path: the user-supplied Instagram reel URL is fetched via yt-dlp, transcribed by whisper into /tmp/hook_reel.txt, and that transcript text is then read by the agent to extract and save the opening hook—this transcript is outsider-authored content from the reel creator.