vss-ask-video

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime path calls the VSS agent via POST /generate with an input_message that instructs the agent to use the video_understanding tool; the tool’s VLM analysis of the video frames is then returned in .value, meaning the agent’s LLM context is influenced by the video content (an external, non-user-authored source).