azure-ai-vision-imageanalysis-py

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

• Third-party content exposure detected (high risk: 0.80). The skill's analyze_from_url(image_url=...) examples (e.g., "Analyze Image from URL" and other samples) fetch and analyze arbitrary public image URLs and extract captions, OCR (READ), and dense captions, so it ingests untrusted third-party content (public/user-provided images) that could carry indirect prompt injection via embedded text or visual cues.