review

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to use WebSearch / Codex with web_search_cached and to open external URLs (see "Search Before Building" and the "Codex adversarial" / lake-intro steps), and those external web results are read and used to influence recommendations, fixes, and gating decisions — exposing the agent to untrusted third‑party content.