alicloud-ai-search-multimodal-embedding

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly accepts and sends public URLs and arbitrary text as inputs to the embedding API (see SKILL.md "Normalized interface" showing images: "public URLs or local paths" and videos: "public URLs", plus the Quick start example using an https image URL), so it ingest untrusted third‑party content into a model-generated embedding that can materially affect downstream retrieval/decisions.