eachlabs-fashion-ai

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

• Third-party content exposure detected (high risk: 0.90). The skill accepts and ingests arbitrary external URLs (e.g., image_url / image_urls / video_url in the Prediction workflows and image_data_url / images_data_url / default_caption / trigger_phrase in the Training sections) which will be fetched from public third‑party sources and consumed by the agent, exposing it to untrusted user-generated content that could carry indirect prompt injections.