From ux-review
This skill should be used when the user asks to "run a UX review", "review UX", "get user feedback on my app", "conduct usability testing", "create UX personas", "test with user personas", "do a usability audit", "find UX issues", "evaluate usability", or invokes the /ux-review command. Orchestrates a multi-agent UX review workflow with simulated persona research and specialist analysis.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ux-review:ux-review-orchestrationThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Orchestrate a comprehensive, multi-agent UX review workflow that produces actionable findings and a prioritized backlog. The workflow simulates real user research by creating personas, conducting simulated interviews and walkthroughs, and dispatching specialist agents for technical and visual analysis.
Orchestrate a comprehensive, multi-agent UX review workflow that produces actionable findings and a prioritized backlog. The workflow simulates real user research by creating personas, conducting simulated interviews and walkthroughs, and dispatching specialist agents for technical and visual analysis.
The review proceeds through 7 phases. Each phase has clear entry criteria, actions, and outputs.
| Phase | Name | Agent(s) | Output |
|---|---|---|---|
| 1 | Intake Interview | Moderator | .ux-review/interviews/intake.md |
| 2 | Persona Generation | Moderator | .ux-review/personas/*.md |
| 3 | Persona Interviews | persona-interviewer (parallel) | .ux-review/interviews/*.md |
| 4 | App Walkthroughs | persona-interviewer (parallel) | .ux-review/walkthroughs/*.md |
| 5 | Specialist Reviews | technical-ux + visual-ux (parallel) | .ux-review/specialist-reports/*.md |
| 6 | Synthesis | Moderator | .ux-review/summary-report.md |
| 7 | Backlog Creation | Moderator + User | .ux-review/backlog.md |
This workflow uses four co-located agents defined in the plugin's agents/ directory:
ux-review-moderator — Orchestrates all phases, conducts intake, generates personas, synthesizes findingspersona-interviewer — Simulates a specific user persona for interviews and walkthroughs (dispatched once per persona)technical-ux-specialist — Evaluates information architecture, task flows, error handling, accessibilityvisual-ux-specialist — Evaluates visual design quality and creates improvement mockupsDispatch agents using the Agent tool with the appropriate agent name. The moderator dispatches persona-interviewers and specialists; the orchestrating session dispatches the moderator.
The following tools and integrations enhance the review but are not required:
Detect the application type from the provided arguments or project context:
--screenshots-only flag passedSelect the appropriate walkthrough strategy from references/walkthrough-strategies.md based on the detected type.
Conduct a structured interview with the user using AskUserQuestion. Follow the question framework in references/interview-guide.md. Cover: app purpose, target users, key workflows, known pain points, business goals, competitive landscape, success metrics, and review scope.
Save responses to .ux-review/interviews/intake.md. If the user is short on time, use the minimum viable intake (4 questions) from the interview guide.
Generate 3-5 simulated user personas based on intake data. Use the profile structure from references/persona-template.md. Ensure diversity across:
Save each persona to .ux-review/personas/[kebab-name].md. Present personas to the user for review before proceeding.
Dispatch one persona-interviewer agent per persona, in parallel. Each agent receives:
.ux-review/interviews/[persona-name].mdInstruct each agent to role-play as its assigned persona, articulating expectations, concerns, must-haves, and deal-breakers.
Dispatch persona-interviewer agents again (one per persona, parallel). Each agent:
.ux-review/walkthroughs/[persona-name].mdDispatch technical-ux-specialist and visual-ux-specialist agents in parallel. Provide each with:
.ux-review/interviews/intake.md.ux-review/personas/.ux-review/walkthroughs/Direct the technical specialist to evaluate: information architecture, navigation, task flows, error handling, accessibility (WCAG 2.1 AA), pattern consistency, cognitive load.
Direct the visual specialist to evaluate: visual hierarchy, typography, color/contrast, spacing, brand consistency, responsive behavior. Instruct it to create HTML mockups for top improvement suggestions.
Read all artifacts from .ux-review/ and synthesize into a summary report. Use the structure from references/report-template.md:
Save to .ux-review/summary-report.md. Present the executive summary and top findings to the user.
Facilitate interactive backlog creation with the user using AskUserQuestion:
Use the item format from references/backlog-template.md. Save to .ux-review/backlog.md.
Always produce a report even with incomplete data. Mark findings with confidence levels: High (directly observed), Medium (inferred from partial data), Low (inferred from code/docs only).
All artifacts are relative to the project root:
.ux-review/
├── interviews/
│ ├── intake.md
│ ├── [persona-name].md
│ └── ...
├── personas/
│ ├── [persona-name].md
│ └── ...
├── walkthroughs/
│ ├── [persona-name].md
│ └── ...
├── specialist-reports/
│ ├── technical-ux.md
│ └── visual-ux.md
├── mockups/
│ ├── [description].html
│ └── ...
├── summary-report.md
└── backlog.md
references/persona-template.md — Persona profile structure and diversity guidelinesreferences/interview-guide.md — Intake question framework with minimum viable setreferences/walkthrough-strategies.md — Per-app-type interaction and evaluation guidesreferences/report-template.md — Summary report structure with severity guidelinesreferences/backlog-template.md — Backlog item format and collaborative creation processnpx claudepluginhub grillergeek/skills --plugin ux-reviewUsability review — evaluate an existing flow or UI against usability heuristics, flag friction points, and recommend fixes. Use when asked to "review the UX", "usability audit", "what's wrong with this flow", "UX feedback", "critique this design", or "why are users dropping off here".
Runs UX regression testing on running apps: static checks for AI slop, accessibility, Nielsen heuristics; dynamic pixel-only user journey validation. Outputs dashboard of UX issues and UI risk signals with suggestions. Use after building or UI changes.
Runs a repo-driven UX audit with journey mapping, heuristic evaluation, and pain-point detection. Supports light (summary) and full (multi-file) modes.