What it does
Hands a finished session to a fresh model that never saw it and asks how well the loop actually ran: did it follow the process, honor the conventions, write real tests, ask at the right moments. Then it routes each finding to a durable fix.
Run it after a run to catch what you didn't notice live.
- Digests the whole transcript: prompts, corrections, the tool timeline, the diff.
- Judges the run against your standards, not its own author's optimism.
- Routes findings back to add-rule or a process note.
Install and run
just this skill
git clone --depth 1 https://github.com/uiverify/uiverifycp -r uiverify/packages/skills/skills/evaluate-run .claude/skills/rm -rf uiverify# then invoke it
/evaluate-runOr install the whole bundle
Claude Code
- 1
/plugin marketplace add uiverify/uiverify - 2
/plugin install uiverify@uiverify
Recommended. Run each command on its own - Claude Code opens a prompt after the first, so pasting both at once fails.
View on GitHub
UI Verify
Built by the team behind UI Verify
Automate this skill end to end
UI Verify captures every story on every pull request and lets an AI judge, or your agent, review the diffs. See how it works.