pubmed-literature-search
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by retrieving and processing unstructured text (abstracts and metadata) from an external database (PubMed).
- Ingestion points: External data is ingested via the NCBI E-utilities API (ESearch, EFetch, ESummary).
- Boundary markers: The skill documentation does not define specific boundary markers or instruction-ignoring delimiters for the retrieved literature content.
- Capability inventory: The skill retrieves metadata and full-text links for agent-side summarization and research analysis.
- Sanitization: No explicit sanitization or filtering of the retrieved publication content is specified in the skill definition.
Audit Metadata