ai-loop

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill follows secure development practices by enforcing a strict 'Spec-Build-Review' cycle. It explicitly instructs the agent to avoid adding extra features, to use human approval gates for destructive or external actions, and to run generated code in sandboxed environments. No signs of prompt injection, data exfiltration, or obfuscation were found in the instructions or examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 01:47 PM
Security Audit — agent-trust-hub — ai-loop