UI Verify
Official@uiverify · United Kingdom
Visual testing for agents - catch unintended UI changes on every PR.
Agent Skills by UI Verify
Showing 15 vetted skills indexed across 1 GitHub repositories.
e2e-verify
Starts a dev stack, authenticates Playwright with the user's browser session, and tests branch changes end-to-end.
evaluate-run
Grade a completed agent run against conventions using a compressed transcript digest and diff.
economical-visual-tests
Consolidates component variants and states into gallery stories to reduce billed visual snapshots.
setup-visual-testing
Set up UI Verify visual regression testing with capture path detection and CI wiring.
storybook-visual-testing
Make Storybook stories deterministic for UI Verify visual regression captures.
vitest-visual-testing
Eliminates flaky diffs in Vitest browser-mode visual tests by freezing data, clocks, and animations.
review-loop
Iterates multi-model code review and fix cycles until the working tree diff is clean.
playwright-visual-testing
Diagnose and fix flaky visual diffs in Playwright archive-replay captures by pinning run-to-run page variation.
babysit-pr
Opens a pull request and polls CI until all checks pass, fixing failures along the way.
making-ui-changes
Guides UI changes through reuse, state coverage, blast-radius checks, and visual test verification.
add-rule
Convert code corrections into conventions-doc rules with enforced lint guards.
check-visual-changes
Renders scoped UI components on the UI Verify fleet and diffs them against the CI baseline.
triage-visual-changes
Triage UI Verify visual regression builds over MCP and accept baselines in bulk.
factory
Implements a requested change and drives it to a merge-ready PR through review, e2e verification, and CI monitoring.
review-multi-model
Reviews git diffs with parallel Claude and Codex reviewers and merges verified findings.
Frequently Asked Questions About UI Verify
FAQPage SchemaWhat tasks can I accomplish with UI Verify's skills?▼
You can set up visual regression testing end to end, author economical snapshot stories, make Storybook/Playwright/Vitest captures deterministic, check visual impact mid-task with uiverify check, triage build diffs via the UI Verify MCP, run multi-model code reviews, babysit PRs to green, and convert corrections into durable lint rules.
Who are UI Verify's skills designed for?▼
They target front-end engineers and coding-agent operators who ship UI changes and need visual regression coverage on every PR. Teams using Storybook, Playwright, Vitest browser mode, or per-snapshot platforms like Chromatic and Percy benefit from the deterministic-capture and snapshot-economy guidance.
How does the practical runtime flow work?▼
The setup-visual-testing skill detects or scaffolds a capture path, installs the uiverify CLI, and wires it into CI. During development, check-visual-changes renders touched components against the real CI baseline, while e2e-verify starts the dev stack on isolated ports and authenticates Playwright with the user's browser session.
What are the prerequisites and dependencies?▼
You need a repository with a capture source: Storybook stories, a Playwright suite, Vitest browser-mode component tests, or raw screenshots. The uiverify CLI must be installed and wired into CI, and build triage requires access to the UI Verify MCP from your coding agent.
How does UI Verify control visual testing costs?▼
The economical-visual-tests skill teaches gallery and matrix stories plus data-driven states so full variant coverage fits in the fewest billable snapshots. This applies to any per-snapshot platform, including UI Verify, Chromatic, Percy, and Playwright screenshots, reducing both billing and diff noise.