uxaudit plugin
A Claude Code plugin for UX regression testing that walks user journeys in a live app and returns a ranked fix plan.
- 59 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The agent opens the app in a Playwright browser and walks journeys, including sign-up.
- The plugin downloads about 300 MB of Chromium on first install.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/uxauditDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent opens the app in a Playwright browser and walks journeys, including sign-up.
- The plugin downloads about 300 MB of Chromium on first install.
This is third-party code. Review the repository files before installing.
What it does
uxaudit reads the project (code, routes, docs, tests) and drafts 5 to 8 journeys for your product. The agent then walks each one in the running app: can someone sign up, reach first value and recover after an error. About 40 more checks run in five groups: traces of generated UI, WCAG 2.2 AA accessibility, usability, core experience and desirability. Each check cites a published source such as WCAG, NN/g, Nielsen heuristics, Baymard or ISO 9241. The result is a prioritized fix list with screenshots that can go straight back to a coding agent. A dashboard keeps run history and comparisons. Invoke it with `/uxaudit:uxaudit my-app`.
Who it is for. Teams shipping agent-built interfaces who want to check that a flow makes sense once automated tests pass.
Good fit when
- End-to-end tests pass, but it is unclear whether a new user reaches first value.
- Before a merge you need accessibility and error-recovery checks.
- An agent built the interface and you want to find typical traces of generated design.
- You need a fix plan with screenshots to hand to the agent.
Not a fit when
- You need speed and Core Web Vitals measurements.
- There is no running app to audit.
- You use Codex or another agent: the README says these hosts are not supported yet.
Example request
Run uxaudit for the my-app app, skip the core-experience check and show the top five fixes with screenshots.Limitations
The README marks the project as experimental: interfaces, checks and outputs may change without notice. It needs Claude Code with plugin marketplace support, and the Max plan is recommended because a full audit takes 30 to 60 minutes. The first install downloads Playwright Chromium, about 300 MB. Only Claude Code is supported, other hosts are planned. Deeper evaluation of decisions and error recovery is still being built.
How to disable. Remove the plugin with `claude plugin uninstall uxaudit@gotalab-uxaudit` and the `gotalab-uxaudit` marketplace from `/plugin`. The downloaded Chromium directory is removed separately.
FAQ
How is uxaudit different from Playwright or Cypress?
They check that a flow executes. uxaudit checks that the flow makes sense to a person: dead ends, weak error recovery, decision cost.
How long does an audit take?
A full audit with journey evaluation takes 30 to 60 minutes. A quick catalog-only check runs with `--skip core-experience`.
Where do the journeys come from?
The agent reads code, routes, docs and tests, generates 5 to 8 journeys for the product and walks them in the live app.
Which check groups exist?
Five: ai-slop, accessibility, usability, core experience and desirability. Every check cites a published source.
Does it work with Codex?
No, the README says Codex and other hosts are not supported yet. The `skills/uxaudit/` bundle is being designed to be portable.
Discussion
No questions or reviews yet.
Related
PluginsA six-skill accessibility plugin: page review, WCAG 2.2 audit with Playwright, audit planning and an improvement roadmap.
A Claude Code plugin for SwiftUI that walks an app like a user and finds unreachable screens, dead ends and placeholder data.
Rails Audit by thoughtbot
Rails Audit (thoughtbot)
A thoughtbot skill for reviewing Rails applications, finding code and test issues and ranking findings by severity
MCP Accessibility Scanner runs axe checks, exercises keyboard focus and saves accessibility reports for web pages
CLI installs a verify-by-running harness into Claude Code and Codex: build and tests, diff review and a separate QA agent.
show-your-work
claude-show-your-work
A plugin that shows images and video in the Claude Code terminal, records Playwright checks and compares before and after
README in short
The plugin checks UX through journeys in a live app and adds about 40 sourced checks. The output is a fix plan with screenshots and a history dashboard.