gnomad-database

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions do not contain any prompt injection patterns, bypass markers, or attempts to override agent behavior. The language is purely instructional and focused on genomic data analysis.\n- [SAFE]: All network operations are directed towards the official gnomAD API domain (gnomad.broadinstitute.org). This is a well-known scientific service, and no sensitive local data access or exfiltration patterns were detected.\n- [SAFE]: The provided Python code snippets use standard libraries (requests, pandas) for data retrieval and processing. There is no evidence of obfuscation, remote code execution (e.g., curl-to-shell), or dangerous dynamic execution patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 11:30 PM
Security Audit — agent-trust-hub — gnomad-database