tw-edu-anti-ai-assessment

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

• Third-party content exposure detected (high risk: 1.00). The skill's SKILL.md MCP connector explicitly lists "WebSearch" (under "MCP 連接器 / Claude Code") instructing the agent to search the open web for "最新抗 AI 評量設計研究" and examples, meaning the agent is expected to fetch and interpret public third‑party webpages whose content could materially affect its analysis and report generation.