karpathy-agentic-guidelines-full

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of natural language instructions (SKILL.md) and an evaluation report (eval-summary-2026-04-17.md). No executable scripts, binary files, or configuration files are present.
  • [SAFE]: No network operations, sensitive file access, or data exfiltration patterns were detected. The skill does not use any tools or external APIs.
  • [SAFE]: No prompt injection, obfuscation, or persistence mechanisms were found. The instructions are focused on software engineering best practices for AI-human collaboration.
  • [SAFE]: While the skill describes an evaluation environment (using local runners and isolated directories), these are part of a descriptive report and do not represent commands executed by the skill itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 05:52 PM
Security Audit — agent-trust-hub — karpathy-agentic-guidelines-full