marketing-psychology

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can be injected via the runtime user-supplied materialText field (and/or text extracted from materialUrl/uploaded materialFile), which the skill’s context composer feeds into the LLM as “粘贴需要分析的页面、报价、活动或漏斗内容”.