tw-edu-exam-generator

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

• Third-party content exposure detected (high risk: 0.90). This skill's SKILL.md explicitly enables a WebSearch connector ("WebSearch：搜尋最新時事素材作為情境題來源（自動啟用）") meaning the agent will fetch and use open/public web content (untrusted third‑party news/material) as a source for generating exam scenarios, so those external pages could indirectly inject instructions that affect generation behavior.