All skills
simota avatar

/vision

@c805268
by shingo imotasimota/agent-skills85 stars
15

Directing UI/UX creative work — redesigns, new designs, trend application, Design System construction, Muse/Palette/Flow/Forge orchestration. Use for design direction. Offers a co-design pair mode.

Use this Skill: https://skilld.dev/gh/simota/agent-skills/vision

This session only. Nothing lands on disk.

referencedesign-system-anti-patterns.md

≈904 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Design System Anti-Patterns

Purpose: Use this file when defining token architecture, screening design-system debt, or setting governance rules.

Contents:

  • DS-01 to DS-08
  • token architecture
  • theme strategy
  • governance rules

Anti-pattern Catalog

ID Anti-pattern Risk Response
DS-01 Token sprawl consistency collapse use a Reference -> Semantic -> Component model
DS-02 Appearance-based naming renaming pain on brand change use semantic naming
DS-03 Hardcoded values token drift lint and CI checks
DS-04 Redundant tokens confusion and maintenance cost regular token audits
DS-05 Design/code naming mismatch team misalignment shared vocabulary
DS-06 Abrupt token deletion downstream breakage staged deprecation
DS-07 Theme-blind token model dark-mode or multi-brand rewrite theme-aware semantics
DS-08 Missing rationale and docs misuse and drift usage notes required

Token Architecture

Layer Role Naming guidance
Reference raw values only category.value
Semantic intent and theme mapping category.purpose.state
Component component-specific bindings component.property.variant.state

Good examples:

  • color.primary.default
  • color.primary.hover
  • text.heading.lg
  • space.layout.section

Bad examples:

  • color.blue
  • --color1
  • button-blue-big
  • primary

Theme Strategy

  • Keep components theme-agnostic.
  • Switch themes by remapping semantic tokens to reference tokens.
  • Support at least:
    • light/dark themes
    • high-contrast theme where required
    • reduced-motion handling

Governance Rules

  • New tokens must declare which layer they belong to.
  • Token removals must follow deprecation: warning -> alternative -> removal.
  • Detect hardcoded values in CI.
  • Keep Figma Variables and code tokens aligned.
  • Review DS-01 to DS-08 during REVIEW mode.

AI-Readable Token Guidelines (2025-2026)

Tokens must carry semantic intent so AI tools make context-aware decisions:

Quality Example Problem
Good color-feedback-error AI understands intent
Good spacing-component-button-padding AI understands context
Bad color-red-500 AI cannot infer purpose
Bad spacing-4 AI cannot infer usage

Additional anti-patterns:

ID Name Symptom Fix
DS-09 AI-unreadable tokens AI generates wrong values Add intent context to token names
DS-10 Static design system (no MCP) AI agents lack real-time token/component data Integrate Figma MCP server for live token sync
DS-11 Token/code naming divergence AI suggestions break design system rules Supply structured rules file to AI agents

Dark Mode Token Integration (2025-2026)

Semantic tokens must be theme-agnostic. Mode switching remaps semantic tokens to reference tokens.

Required semantic token categories:

  • color.background.surface (not color.white / color.gray-900)
  • color.text.primary / color.text.secondary
  • color.border.default
  • color.feedback.error / warning / success / info
  • color.interactive.primary / hover / active / disabled

CSS light-dark() (modern approach):

:root { color-scheme: light dark; }
.surface { background: light-dark(#ffffff, #111111); }

Anti-pattern DS-07 (Theme-blind token model) is the most common cause of dark-mode rewrites. Adopt semantic tokens from day one.

Source: SKILL.md on GitHub

No alerts14d5 checks · Risk SAFE
  • Gen Agent Trust Hub14d

    The 'vision' skill is a comprehensive design-direction tool that operates by defining strategies and delegating work to other specialized agents. It contains robust internal guardrails, including mandatory accessibility standards (WCAG 2.2 AA) and ethical AI principles. No security risks were identified.

  • Socket14d

    No alerts

  • Snyk14d

    Risk: LOW · No issues

  • Runlayer6mo

    9 files scanned · No issues

  • ZeroLeaks5mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at c805268. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated 2 weeks ago

README badge

README badge for simota/agent-skills/vision