ai-talking-head

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly shows workflows that accept and fetch arbitrary external URLs for source videos and audio (e.g., the Lip-Sync Overlay and Handoff examples with "video": "https://... (source video URL)" and "audio": "https://... (audio file URL)"), meaning the agent ingests untrusted third‑party/user-generated content which can materially affect generation and subsequent actions.