pubmed-database
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEOBFUSCATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The technical support email address listed in the support resources section ('vog.hin.mln.ibcn@seitilitue') is provided in a reversed format. When read backwards, it reveals the address 'eutilities@ncbi.nlm.nih.gov'.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from an external source, creating a potential vector for indirect prompt injection.
- Ingestion points: The skill retrieves search results, abstracts, and full-text metadata from the NCBI E-utilities API (eutils.ncbi.nlm.nih.gov).
- Boundary markers: There are no instructions to the agent to isolate API responses with delimiters or to ignore any instructions embedded in the retrieved scientific abstracts.
- Capability inventory: The skill uses the Python 'requests' library to perform network operations and provides templates for parsing and displaying external article content.
- Sanitization: No mechanisms are described for sanitizing or filtering the content fetched from the API before it is processed by the agent.
Audit Metadata