verification-before-completion

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional and contains no executable code, network operations, or external dependencies. It defines a set of behavioral constraints intended to improve task reliability and prevent deceptive reporting by the agent by requiring empirical evidence (such as test results or build logs) before the agent can claim a task is completed.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 09:48 AM
Security Audit — agent-trust-hub — verification-before-completion