updating-knowledge

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly requires using web_search and web_fetch (see "Preflight Check" and "Tool Usage Notes" which state web_fetch "can fetch ANY URL") and mandates gathering and synthesizing 3–5 diverse sources including community discussions, so the agent will ingest untrusted public web content and use it to influence decisions.