prompt-engineering

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

• Third-party content exposure detected (high risk: 0.90). The skill's templates and examples explicitly instruct the agent to perform web searches and handle search observations (e.g., ReAct SEARCH actions in references/react.md, the web_search/tool in references/json-format.md, and prompt-chaining steps that call "web search" or "Search Knowledge Base" in references/prompt-chaining.md), which means the agent is expected to fetch and read open/public third‑party content that may be untrusted.