ai-tutorials

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

• Third-party content exposure detected (high risk: 0.90). The skill's Step 10 MCP lesson and its project task "全能情报员" explicitly instruct configuring a Web Search MCP (e.g., brave-search) so the agent autonomously searches and ingests public web content (latest资讯) and compiles it into local Markdown, meaning the agent will read and act on untrusted third‑party content.