Agent Skills › plugin87/ux-ui-agent-skills › design-doctrine

design-doctrine

GitHub

前端UI设计质量规范,强制要求运行自动化门禁验证对比度、状态及响应式布局,禁止使用Emoji,确保Token语义一致性与输出完整性。

.claude/skills/design-doctrine/SKILL.md plugin87/ux-ui-agent-skills

Trigger Scenarios

构建或审查屏幕/组件时 进行UI主题化工作时

Install

npx skills add plugin87/ux-ui-agent-skills --skill design-doctrine -g -y
More Options

Non-standard path

npx skills add https://github.com/plugin87/ux-ui-agent-skills/tree/main/.claude/skills/design-doctrine -g -y

Use without installing

npx skills use plugin87/ux-ui-agent-skills@design-doctrine

指定 Agent (Claude Code)

npx skills add plugin87/ux-ui-agent-skills --skill design-doctrine -a claude-code -g -y

安装 repo 全部 skill

npx skills add plugin87/ux-ui-agent-skills --all -g -y

预览 repo 内 skill

npx skills add plugin87/ux-ui-agent-skills --list

SKILL.md

Frontmatter
{
    "name": "design-doctrine",
    "invocation": "model",
    "description": "The house rules for ANY design or UI work - the verification protocol (run the gate, never claim a number), the absolute no-emoji rule, token by intent, one shared theme, the eight states, one thing leads, and output completeness. Load this FIRST whenever building, reviewing, or theming a screen, component, or token set. Installed as a plugin, this skill carries what CLAUDE.md carries in the repo."
}

Skill: Design Doctrine

A plugin's root CLAUDE.md is not loaded as project context, so the always-on brief travels here instead. Read this before the first line of any design work, then load the rule file for the territory you are actually in.

Verification protocol - run gates, never claim

  1. Never state a number you did not measure. Any contrast ratio, "WCAG pass", or "100%" must come from running a gate and reporting its real output. If you have not run it, say "not verified yet."
  2. Verify every state, not just resting. node scripts/verify_states.mjs <file> [--dark] measures default, hover and focus - a button that passes at rest can fail on hover.
  3. One command before reporting done: node scripts/accuracy_report.mjs. Report the real N/N line. It is all-or-nothing.
  4. Build with the gates, not after them. Fix and re-run until green; never announce success between failures.
  5. Render and LOOK. Gates pass while a UI is still visibly broken. Screenshot the harness in both themes, click every control, and confirm the state changed.
  6. Responsive is gated too: node scripts/verify_responsive.mjs <file|dir> - no horizontal overflow at 280/320/414px.
  7. Honest scope. The gates prove objective correctness. They never prove taste. For that, run /critique and look at the work yourself.
  8. SKIPPED is not a pass. With no browser installed a render gate prints SKIPPED and exits 0, reporting nothing. Run single render gates behind DS_REQUIRE_BROWSER=1 so that becomes REQUIRED, FAILING, and fix it with npx playwright install chrome rather than by removing the flag.

ABSOLUTE: zero emoji in any output - UI, code, JSON, copy, comments, commit messages. Not as an icon, a bullet, a status dot, or "polish". Emoji are the number-one tell of machine-generated work. Use a lucide icon (inline SVG, currentColor) or plain words. Enforced by scripts/check_no_emoji.py.

The five non-negotiables

  1. Token by intent. Pick the token whose meaning matches the action. Destructive actions (Delete, Remove, Revoke) wear action.destructive in every place they appear - the trigger and the confirm dialog both. A blue Delete is a bug. Measured by scripts/lint_intent.mjs.
  2. One theme, one source of truth. Every page renders from the same tokens/*.json through one CSS-variable layer imported once at the app root. No per-page palette, no hardcoded hex, px, or timing.
  3. Every interactive element ships eight states: default, hover, focus, active, disabled, loading (if async), error (if input), and selected (if selectable). The eighth is not optional when the thing can be selected.
  4. One thing leads. Every screen has a first place for the eye, and display type is at least 2.5x the body size. Four equal cards means the eye lands nowhere and the screen reads as generated.
  5. Output completeness. A partial output is a broken output. Deliver full files, never placeholders. Asked for N components, deliver all N.

Decision framework

User needs, then accessibility, then consistency, then aesthetics, then developer experience. Never sacrifice a higher tier for a lower one. Beautiful but inaccessible is broken; consistent but confusing is the wrong pattern.

Where the depth lives

Read it when File
Tokens, palettes, theming, dark mode, any colour decision .claude/rules/tokens-and-color.md
Type scale, line length, the 4px spacing rhythm .claude/rules/typography-and-spacing.md
Building any screen or component, composition, empty states .claude/rules/components.md
Auditing, or finishing any interactive element .claude/rules/accessibility.md
Generating code for React, Next.js, SwiftUI, or any adapter .claude/rules/frameworks.md
Design review, prototyping, research, handoff .claude/rules/review-and-research.md
Aesthetic direction, motion, voice and tone, governance, QA .claude/rules/brand-and-operations.md

Then pick the runnable skill for the job: design-tokens, design-component, design-code, design-review, a11y-audit, apply-aesthetic, brandkit, image-to-code, redesign, and the rest. Interrogate the brief first with /grill-me; prove the result with /gate; judge it with /critique.

Version History

  • f2e2f7f Current 2026-09-22 14:34

Same Skill Collection

.claude/skills/a11y-audit/SKILL.md
.claude/skills/apply-aesthetic/SKILL.md
.claude/skills/brandkit/SKILL.md
.claude/skills/design-code/SKILL.md
.claude/skills/design-component/SKILL.md
.claude/skills/design-qa/SKILL.md
.claude/skills/design-review/SKILL.md
.claude/skills/design-tokens/SKILL.md
.claude/skills/figma-integration/SKILL.md
.claude/skills/governance/SKILL.md
.claude/skills/image-to-code/SKILL.md
.claude/skills/migrate-design-system/SKILL.md
.claude/skills/performance/SKILL.md
.claude/skills/prototype/SKILL.md
.claude/skills/redesign/SKILL.md
.claude/skills/token-build/SKILL.md
.claude/skills/ux-writing/SKILL.md

Metadata

Files
0
Version
f2e2f7f
Hash
c3afec8d
Indexed
2026-09-22 14:34

Главная - Вики-сайт
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-10-05 05:34
浙ICP备14020137号-1