munger-perspective

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md explicitly mandates live WebSearch/tool use for "芒格式研究" ("⚠️ 必须使用工具（WebSearch等）获取真实信息，不可跳过." in Step 2), which requires the agent to fetch and read public third‑party web content (news, webpages, etc.) and to base its judgments on that untrusted external material—creating a clear avenue for indirect prompt injection.