tc-action

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

• Third-party content exposure detected (high risk: 0.70). The required workflow in SKILL.md instructs the agent to use “联网、浏览器或公开数据库工具” and run external searches at runtime (public web/content/database sources), which can ingest outsider-authored free text into the LLM context via fetched pages/search results.