lynx-web-search

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly uses lynx to fetch and dump open web pages and search engine results (see the "Engine URL templates" including google/reddit/github and the lynx -dump "$url" / lynx -source "$url" commands), so the agent will read untrusted, user-generated third‑party content and could be exposed to indirect prompt injection.