ai-video-generation

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly shows runtime inputs that fetch and ingest arbitrary external URLs (e.g., "image_url", "audio_url", "video_url" examples and the media-merger "videos" array) so the agent will read untrusted public content from third-party URLs as part of its workflow, which could allow indirect prompt-injection via media/audio/text embedded in those resources.