designing-ux

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

• Third-party content exposure detected (high risk: 0.80). The skill's instructions (e.g., references/AIX-COMPUTATION-UX.md, references/AGENTIC-DESIGN.md and COPILOT-PATTERNS.md) explicitly describe using external searches/RAG, web/tool/plugin access and browser/computer use—i.e., fetching and ingesting public third‑party content (web pages, plugins, external data) that the agent is expected to read and use to drive decisions—so untrusted third‑party content can influence tool actions.