llm-scientific-discovery-guide

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

• Third-party content exposure detected (high risk: 0.80). The skill's HypothesisGenerator example explicitly lists external knowledge_sources=["pubmed", "openalex"] and the workflow describes literature mining and evidence-based hypotheses, so the agent would ingest and act on public third-party content from those sites as part of its workflow.