video-content-reviewer

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

• Third-party content exposure detected (high risk: 0.80). The skill's Information Gathering section explicitly instructs the agent to "work with video files, links, transcripts, or text descriptions" and to ask the user to "paste a link," meaning the agent will ingest and interpret arbitrary public creator videos/transcripts (e.g., social posts, YouTube/TikTok) as part of its required review workflow, which could allow indirect prompt injection from untrusted third‑party content.