verification-before-completion

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a procedural framework for agent reliability. It mandates that claims of success must be supported by fresh verification data (command outputs), reducing the likelihood of hallucinations or false success reports.
  • [SAFE]: No malicious command execution, data exfiltration, or credential harvesting patterns are present. The skill focuses on standard software development verification steps (tests, builds, linting).
  • [SAFE]: The skill does not download external content, use obfuscation, or contain prompt injection attempts. It serves as a safety/reliability enhancement for the agent's workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 02:36 AM
Security Audit — agent-trust-hub — verification-before-completion