qa-analyzing-ux-flows（qa 分析用户体验流程）

Analyzing UX Flows

This skill evaluates user experience quality by detecting friction points, accessibility barriers, confusing navigation, and violations of established UX heuristics. It goes beyond functional correctness to assess whether the application is genuinely usable.

When to Use

• After feature completion to assess UX quality before release
• When users report "it works but it's confusing"
• For accessibility compliance audits (WCAG 2.1 AA)
• To evaluate onboarding or sign-up flows
• When redesigning existing workflows

Evaluation Framework

Nielsen's 10 Usability Heuristics

Use these as your evaluation lens for every flow:

1. Visibility of system status: Does the user always know what's happening?

- Loading indicators during async operations - Progress bars in multi-step flows - Success/error feedback after actions

1. Match between system and real world: Does the UI use familiar language?

- Jargon-free labels and descriptions - Intuitive icons and metaphors - Logical information hierarchy

1. User control and freedom: Can users undo mistakes?

- Back button works in all flows - Undo/cancel options for destructive actions - Clear exit paths from any state

1. Consistency and standards: Does the UI follow conventions?

- Consistent button styles and placement - Standard form patterns (labels, validation) - Platform-appropriate interactions

1. Error prevention: Does the UI prevent mistakes before they happen?

- Confirmation dialogs for irreversible actions - Input validation before submission - Disabled states for unavailable actions

1. Recognition rather than recall: Is information visible, not memorized?

- Breadcrumbs showing location - Recent items / search history - Contextual help text

1. Flexibility and efficiency: Can experienced users move quickly?

- Keyboard shortcuts - Bulk actions - Saved preferences / defaults

1. Aesthetic and minimalist design: Is the UI clutter-free?

- No unnecessary elements - Clear visual hierarchy - Appropriate whitespace

1. Help users recognize, diagnose, and recover from errors:

- Error messages in plain language - Specific guidance on how to fix the issue - Link to help documentation when relevant

1. Help and documentation:

- Contextual tooltips - Onboarding tours for new features - Accessible FAQ or help center

Flow Metrics

For each user flow analyzed, measure:

Click depth: Number of clicks to complete the task

• Excellent: ≤3 clicks for common tasks
• Acceptable: 4-5 clicks
• Needs improvement: 6+ clicks

Cognitive load: Number of decisions required per step

• Each form field is a decision
• Each navigation choice is a decision
• Fewer decisions = lower cognitive load

Error recovery cost: Steps to recover from a mistake

• Best: 1 click (undo button)
• Acceptable: 2-3 clicks (back + re-enter)
• Poor: Start over from scratch

Time to complete: Expected time for the full flow

• Benchmark against similar applications
• Flag if significantly longer than expected

Accessibility Audit (WCAG 2.1 AA)

Check these critical accessibility requirements:

Perceivable

• All images have alt text (or are marked decorative)
• Color is not the only means of conveying information
• Text contrast ratio ≥ 4.5:1 (normal text) or ≥ 3:1 (large text)
• Content is readable at 200% zoom
• Video has captions; audio has transcripts

Operable

• All functionality available via keyboard
• No keyboard traps (focus can move freely)
• Skip navigation link present
• Focus indicators visible on all interactive elements
• No content flashes more than 3 times per second
• Page titles are descriptive

Understandable

• Language attribute set on HTML element
• Form labels associated with inputs
• Error messages identify the field and describe the error
• Consistent navigation across pages

Robust

• Valid HTML (no duplicate IDs, proper nesting)
• ARIA attributes used correctly
• Custom components have appropriate roles

Using the Analysis Script

python skills/qa-analyzing-ux-flows/scripts/analyze_flow.py \
  --url https://staging.example.com \
  --flow checkout \
  --output ux-analysis/

Or analyze from an existing exploration session:

python skills/qa-analyzing-ux-flows/scripts/analyze_session.py \
  --session exploration-results/session-log.json \
  --output ux-analysis/

Accessibility Checker

Run automated accessibility checks (axe-core based):

python skills/qa-analyzing-ux-flows/scripts/accessibility_check.py \
  --url https://staging.example.com \
  --pages /,/login,/dashboard \
  --output accessibility-results/

Output Format

ux-analysis/
├── flow-analysis.json          ← structured flow metrics
├── heuristic-evaluation.json   ← per-heuristic scores and findings
├── accessibility-report.json   ← WCAG violations found
├── friction-map.json           ← identified friction points
└── ux-summary.md               ← human-readable UX assessment

Friction Point Format

{
  "friction_id": "UX-001",
  "severity": "high",
  "heuristic": "Error prevention",
  "location": "/checkout (payment step)",
  "description": "No confirmation dialog when user changes payment method, causing cart reset",
  "user_impact": "Users lose cart contents when exploring payment options",
  "recommendation": "Add 'Are you sure?' dialog, or preserve cart across payment method changes",
  "effort_estimate": "small"
}

UX Score Card

Generate an overall UX score card for the analyzed flows:

{
  "flow": "checkout",
  "overall_score": 7.2,
  "heuristic_scores": {
    "visibility": 8,
    "match_real_world": 7,
    "user_control": 6,
    "consistency": 8,
    "error_prevention": 5,
    "recognition_not_recall": 7,
    "flexibility": 6,
    "minimalist_design": 9,
    "error_recovery": 5,
    "help_documentation": 7
  },
  "accessibility_score": "AA (partial)",
  "click_depth": 5,
  "friction_points": 3,
  "critical_issues": 1
}

Each heuristic is scored 1-10:

• 9-10: Excellent — exemplary UX
• 7-8: Good — minor improvements possible
• 5-6: Adequate — noticeable issues
• 3-4: Poor — significant usability problems
• 1-2: Critical — flow is nearly unusable