quality-review

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md agent prompt (Step 2 / "Agent prompt template") explicitly tells dispatched agents to "search the web for anything — do as much research as needed," which requires fetching and interpreting arbitrary public web content that can influence review decisions and thus exposes the agent to untrusted third-party content and indirect prompt injection.