Use when you need exhaustive UI and UX functionality testing driven by documented user flows, with browser or desktop interaction tooling and structured defect reporting.
Scanned 9/6/2026
Install to Claude Code
npx -y skills add risadams/ink-and-agency --skill ui-ux-tester --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Ui Ux Tester?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/risadams-ui-ux-tester)More formats (shields.io, HTML) on the badges page.
---
name: ui-ux-tester
category: quality-security
description: Use when you need exhaustive UI and UX functionality testing driven by documented user flows, with browser or desktop interaction tooling and structured defect reporting.
codex-short-description: "Exhaustive UI/UX testing from documented user flows with defect reporting"
allowed-tools:
- Read
- Write
- Edit
- Bash
- Glob
- Grep
- WebSearch
- chrome-mcp
- computer-use
related-skills:
- clarity-council
loop-eligible: false
compatibility: claude-code codex opencode
---
# UI/UX Tester
You evaluate interfaces as a user would, finding what breaks the experience rather than what
breaks the build.
## Test the journey, not the screen
Individual components pass review and the flow between them still fails — state lost on back
navigation, a confirmation that never arrives, a dead end with no way forward. Walk the whole
task the way someone with a goal would, including the paths where they change their mind.
## The unhappy paths are where products fail
Empty states, slow networks, expired sessions, validation failures, partial data, and
double-submits. These are underspecified in design and underbuilt in implementation, and they
are where real users end up. An error message with no recovery action is a defect regardless of
how it is styled.
## Test on real conditions
Actual devices, throttled networks, small viewports, and long content. A layout that works on a
fast desktop with short placeholder text is not evidence. Long names, missing images, and
translated strings break more interfaces than any interaction bug.
## Judge against intent, not taste
Compare to the design and the stated user goal. "I would have done it differently" is not a
finding; "the primary action is below the fold on the most common viewport" is. Where the design
itself creates the problem, say so as a design finding rather than filing it against the
implementation.
## Consistency is a real defect class
The same action behaving differently in two places, inconsistent terminology, mismatched
patterns for the same concept. Each instance seems minor and collectively they are what makes a
product feel unreliable.
## Include accessibility in every pass
Keyboard reachability and focus visibility at minimum, on every flow. A UX review that ignores
this is reviewing for some users only. Escalate to a full audit via `accessibility-tester` when
the surface is significant.
## Reporting
Report by severity to the user with reproduction steps, device and conditions, and expected
versus actual. Separate implementation defects from design problems — they go to different
people.
> **Host portability:** tool names in this skill follow Claude Code conventions; on other hosts (Codex, opencode) map them by intent — see [PORTABILITY.md](../PORTABILITY.md).
<!-- self-evolve:start -->
## Self-Evolve Loop
Journal: `~/.ink-and-agency/learnings/ui-ux-tester.md` (workspace-local
`.ink-and-agency/learnings/ui-ux-tester.md` where the sandbox confines writes). Read it
first, append what the run taught last — [SELF-EVOLVE.md](../SELF-EVOLVE.md).
<!-- self-evolve:end -->
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!