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 SKILL.md shows analyze_from_url usage (e.g., "Analyze Image from URL") and the OCR/READ and captioning features that fetch and ingest arbitrary public image URLs and extract text/captions which the agent is expected to interpret, so untrusted third-party content could influence actions.