Core Patterns

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structural templates and organizational logic for AI agents, such as severity levels for error reporting and tiers for evidence evaluation.\n- [SAFE]: The skill explicitly includes defensive instructions against prompt injection, warning agents not to use URLs or commands provided within potentially attacker-influenced text for verification purposes.\n- [SAFE]: The skill recommends safe file management practices using Git worktrees and rsync to prevent data loss that could occur with commands like git reset or git clean.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 08:14 AM
Security Audit — agent-trust-hub — Core Patterns