modeling-activation-metrics

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of markdown documentation and SQL templates (HogQL and dbt). No executable scripts, binaries, or remote dependencies are included.
  • [SAFE]: The skill includes explicit defensive instructions for the AI agent to treat candidate event names as untrusted input (Category 8 mitigation) and to verify definitions with the user before performing persistent actions like view creation.
  • [SAFE]: No obfuscation, hardcoded credentials, data exfiltration patterns, or unauthorized network operations were detected in any of the provided files.
  • [SAFE]: SQL patterns provided are standard data modeling queries and do not contain dangerous operations or injection vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 10:15 PM
Security Audit — agent-trust-hub — modeling-activation-metrics