agent-development

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly defines a web_search tool in the "Tool Design" section (search_tool -> search_api) and the ReAct/Plan-and-Execute workflows ingest tool "Observations" into agent history and decision prompts, so the agent fetches and interprets open web/untrusted content that can influence subsequent actions.