verification-before-completion

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill does not contain any malicious code, external downloads, or data exfiltration patterns. It primarily consists of instructions and best practices for task verification.
  • [PROMPT_INJECTION]: The skill uses strong instructional language to enforce behavioral constraints on the agent. While the tone is authoritative (using terms like "Iron Law" and "Non-negotiable"), the instructions are aimed at improving operational accuracy and do not attempt to bypass core safety guardrails or ignore previous system instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 12:43 PM