NYC

database-sharding

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): No instructions attempting to override agent behavior or bypass safety filters were detected.
  • [Data Exposure & Exfiltration] (SAFE): SQL examples use placeholder domains (e.g., example.com) and do not contain hardcoded credentials, API keys, or sensitive local file paths.
  • [Remote Code Execution] (SAFE): The skill does not download or execute external scripts, and no unverifiable packages are required.
  • [Indirect Prompt Injection] (SAFE): The skill consists of static SQL templates and does not identify any surfaces for untrusted data ingestion that would lead to prompt injection vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:10 PM