verification-before-completion
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists solely of Markdown-based guidelines and does not contain any executable scripts, configuration files, or code.
- [PROMPT_INJECTION]: The skill uses highly imperative and coercive language (e.g., "The Iron Law", "Non-negotiable", "If you lie, you'll be replaced") to mandate verification steps. This is interpreted as forceful behavioral instruction rather than a malicious attempt to bypass safety constraints.
- [SAFE]: No external downloads, credential exposure, or persistence mechanisms were detected.
Audit Metadata