hiring-review

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md workflow (Step 3 "No silent supplement" and related instructions to "search the web — results will be tagged '[web search — verify]'" and "Cite primary sources") explicitly requires the agent to fetch and interpret public web/legal-research sources, so untrusted third‑party content can directly influence the agent's decisions.