prompt-compile

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a structured prompt engineering methodology (Compile-Execute-Verify) designed to increase task deterministic behavior.
  • [SAFE]: No evidence of malicious command execution, data exfiltration, or unauthorized file access was detected.
  • [SAFE]: The skill references a local 'prompt-compile-gate.sh' hook for triggering logic, but does not perform remote downloads or script execution from untrusted sources.
  • [SAFE]: References to reading project files (CLAUDE.md, design tokens, etc.) are standard for agent operations and do not constitute a high-risk data exposure vector within this context.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 02:37 AM
Security Audit — agent-trust-hub — prompt-compile