verify-before-complete

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides procedural instructions to enforce verification rituals. It guides the agent to identify and run appropriate local verification commands (e.g., npm test, pytest, cargo test) before concluding a task.
  • [COMMAND_EXECUTION]: The skill references shell execution tools like async_bash and bg_shell for running project test suites and builds. These are standard operations for a development agent and are used here to ensure work quality rather than for malicious purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 07:10 PM
Security Audit — agent-trust-hub — verify-before-complete