gemini-integration

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

• Third-party content exposure detected (high risk: 0.90). The skill includes src/services/ai/vision.ts's analyzeImageFromUrl which performs fetch(imageUrl) to retrieve arbitrary images from the web and then passes that untrusted third-party content into the Gemini model for analysis, meaning external content can directly influence model outputs and subsequent actions.