hypothesis-generation

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly instructs the agent to fetch and ingest literature from open public sources (e.g., "Use WebFetch with PubMed URLs", "Use WebSearch", and sites like arXiv/bioRxiv/general web) as part of its literature-search and synthesis workflow, exposing it to untrusted third-party content that could carry indirect prompt-injection payloads.