What it does
Runs a multi-model review over your working tree, applies the verified fixes, and re-reviews, because the fix that clears one issue often introduces the next. It stops once a round finds nothing that matters, and sweeps for missing stories, tests and stale comments in one pass instead of one per round.
Run it the moment you finish a change, before you open a PR.
- Two models review the same diff; every finding is verified against the code.
- Loops review, fix, review until a pass finds no real bug or gap; polish never keeps it going.
- A long run tells you which rules are missing, and offers to capture them.
Install and run
just this skill
npx skills add uiverify/uiverify --skill review-loop# then invoke it
/review-loopInstall these alongside it
Review loop runs these skills, so grab them too - or just install the whole bundle below.
Or install the whole bundle
terminal
npx skills add uiverify/uiverify
One command, any agent - Cursor, Codex, Copilot, Claude Code. Run npx skills update to pull the latest.
View on GitHub
