karpathy-guidelines

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of instructional text intended to guide the agent's coding behavior. No executable code or automation scripts are present.
  • [SAFE]: No security threats such as prompt injection, data exfiltration, or obfuscation were identified.
  • [SAFE]: The skill does not access sensitive files, environment variables, or perform network operations.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 08:10 AM
Security Audit — agent-trust-hub — karpathy-guidelines