UI Acceptance Auditor

Goal: turn subjective "looks fine" into measurable acceptance.

SSOT

• •../../ui-acceptance-criteria.md
• •../../AI_UI_Skill_Design_Document.md
• •Frontend development specification document (Chinese source file in repo)
• •../contracts/quality-gates.md

Three Acceptance Levels

1. •Level 0 baseline compliance (Pass/Fail)

• •Syntax, typing, responsiveness, basic interaction, and no blocking errors.
• •Security baseline (input escaping, sensitive-data exposure check) and mainstream browser compatibility.

1. •Level 1 quality standard (0-100)

• •Visual hierarchy, color usage, typography quality, spacing discipline, and component quality.

1. •Level 2 excellence experience (bonus)

• •Micro-interactions, design highlights, responsive details, and enhanced accessibility.

Execution Flow

1. •Run Level 0 first

• •Any fail returns blocking issues immediately; do not continue scoring.

1. •Run Level 1 next

• •Output dimension scores, total score, and lowest-scoring weakness.

1. •Evaluate Level 2 last

• •Treat as enhancement only; it must not mask baseline issues.

1. •Output five-dimension acceptance matrix

• •Function / Visual / Code / Performance / Security & Compatibility.

Recommendation Thresholds

• •>= 85: release-ready (minor polish recommended)
• •70-84: usable, but fix key weaknesses before release
• •< 70: continue iteration before release

Output Contract

1. •Acceptance conclusion (pass / conditional pass / fail)
2. •Scoring table by dimension
3. •Five-dimension matrix (function/visual/code/performance/security-compatibility)
4. •Top 5 issues with fixes (sorted by severity)
5. •Re-test guidance (what to verify again)

Collaboration

• •Final quality gate in ui-codegen-master.
• •Pair with ui-self-reviewer to form external acceptance + internal self-review.