ugc-talking-head-video

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

• Third-party content exposure detected (low risk: 0.10). The workflow prompts the agent to search external social media platforms (Instagram and TikTok/Douyin) for video references using treg catalog search, introducing low-risk exposure to outsider-authored free text from selected search results.

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

• Potentially malicious external URL detected (medium risk: 0.30). The repository link points to a GitHub URL owned by an independent, user-selected community account (agentara), representing an unofficial source for the portrait-clone skill.