Verify Work

This runs systematic verification across unit, integration, E2E, mobile, and security test lanes before you hand off work. It loads your acceptance criteria from whatever planning doc you have, picks the right test types based on what changed, then executes them smallest to largest while capturing evidence like logs and screenshots. The comparative audit step is smart: it proves the before and after states for bug fixes, which actually matters during review. It traces every AC back to your BRD or PRD and blocks instead of guessing when environments are broken. The evidence gets written into a walkthrough doc with pass/fail status per criterion. Use this when you need proof of delivery, not just "it works on my machine."

Install to Claude Code

npx -y skills add hoangnguyen0403/agent-skills-standard --skill verify-work --agent claude-code