pubmed-search
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches citation data and abstracts from the official NCBI PubMed API endpoints at
eutils.ncbi.nlm.nih.gov. This is a well-known scientific service provided by the US National Library of Medicine. - [COMMAND_EXECUTION]: Shell commands are used to perform HTTP requests via
curland to parse the resulting JSON data usingpython3. The python script used is a static inline command for data processing and does not execute remote logic. - [PROMPT_INJECTION]: The skill is subject to indirect prompt injection surfaces as it processes uncontrolled external data (paper titles and abstracts) from the PubMed database.
- Ingestion points: Data retrieved from the
esearchandefetchAPI endpoints. - Boundary markers: Not present; the agent is instructed to report the results exactly as returned by the API.
- Capability inventory: Access to
curlfor network requests andpython3for JSON parsing. - Sanitization: No specific sanitization or filtering is applied to the retrieved biomedical text before presentation to the agent.
Audit Metadata