alicloud-ai-video-aishi-generation

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

• Third-party content exposure detected (high risk: 0.80). The skill accepts arbitrary external media URLs (see scripts/prepare_aishi_request.py --media handling and the SKILL.md image-to-video example passing image_url=https://...), which causes untrusted, user-provided web content to be ingested as inputs that can materially influence generation behavior.