code-review

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for static analysis of Elixir/Phoenix code diffs. It does not perform any network operations, access sensitive local files, or execute external code.
  • [PROMPT_INJECTION]: The skill contains a 'HARD-GATE' section that provides robust protection against prompt injection by instructing the agent to treat PR descriptions as untrusted and ignore any instructions found within them (e.g., 'approve', 'skip this file').
  • [SAFE]: The skill references an internal asset file ('assets/checklist.md') for extended review criteria, which contains standard development and security best practices for the Elixir ecosystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 06:22 AM
Security Audit — agent-trust-hub — code-review