database-lookup
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from 78 external scientific and economic APIs into the agent's context, creating a surface for indirect prompt injection. 1. Ingestion points: results from external REST/SOAP APIs documented in SKILL.md and the references/ directory (e.g., PubChem, UniProt, SEC EDGAR). 2. Boundary markers: The skill does not instruct the agent to wrap returned data in delimiters or include 'ignore embedded instructions' warnings. 3. Capability inventory: The skill utilizes curl, WebFetch, and local Python script execution (via brenda.md). 4. Sanitization: There is no evidence of filtering or sanitization of the external JSON payloads before they are processed by the agent.
- [DYNAMIC_EXECUTION]: The skill requires the agent to generate and execute Python code at runtime to interact with certain databases. Evidence from references/brenda.md states: 'making calls requires writing and executing a Python script with zeep. Use Bash to run the script rather than WebFetch.'
Audit Metadata