deep-dive

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

• Third-party content exposure detected (high risk: 0.90). The skill's Phase 2 research workflow explicitly instructs agents to run WebSearch queries and use WebFetch/WebFetch-like reads on public web results ("Run 5-7 WebSearch queries" and "Use WebFetch on the 5-8 most substantive results"), which ingests untrusted, user-generated or public third‑party webpages that the agent must read and act on for synthesis, enabling indirect prompt injection.