azure-ai-vision-imageanalysis-java

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime path client.analyzeFromUrl(imageUrl, ...) fetches a user-supplied URL and ingests the resulting image content (including OCR text via VisualFeatures.READ) into the agent’s context as readable text, which can include outsider-authored free text from the remote page/image.