academic

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly spawns research/survey agents that fetch and ingest open web content—e.g., the "Paper Crawler" which collects papers from DBLP + OpenAlex APIs and the "Research Analyst" (read + web)—and the workflows ("collect papers on X", "literature survey on X") require reading that third‑party content, so untrusted web content could influence agent decisions.