Color System logo

Color System

Organization
KyaniteLabs
color-system

OKLCH color-system guidance for cohesive palettes and contrast-safe tokens. Use when choosing colors, generating ramps, fixing muddy palettes, creating semantic state colors, checking WCAG contrast, or defining theme variables.

Overview

PublisherKyaniteLabs
Repositorytastecheck
Skill namecolor-system
Stars
10
Forks
7
Bundled files
5
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by KyaniteLabs on GitHub. Read the source before you install it.

Installation

Install the Color System AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/KyaniteLabs/tastecheck/tree/main/skills/color-system
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/KyaniteLabs/tastecheck.git /tmp/tastecheck
mkdir -p .claude/skills
cp -r /tmp/tastecheck/skills/color-system .claude/skills/color-system
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Color System in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Color System on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Color System is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Color System

Build a palette whose roles survive text, controls, themes, and gamut limits. OKLCH makes lightness/chroma easier to reason about than HSL/hex, but does not choose art direction or guarantee contrast; the rendered pair is the test.

Why OKLCH

oklch(L C H): L 0–1 perceptual lightness, C chroma (0 = gray, ~0.37 max real), H 0–360 hue. Similar L values start even ramps, but hue/chroma still need inspection. Adjust one property without wrecking the others. It is P3-ready and degrades to sRGB; provide a hex fallback for very old targets.

The method

  1. Choose structure from evidence. A restrained product may need one hue/neutrals; data or editorial systems may need several. Record each hue's job.
  2. Build a fixed-step lightness ramp: 10–12 stops from ~0.97 (50) to ~0.30 (900). Keep H constant; vary L; let C peak mid-range and taper at extremes.
  3. Choose neutral temperature deliberately. Brand tinting can create cohesion; true neutrals can suit data, photography, or strict status colors.
  4. Pick semantic colors (success/error/warning/info) at matched L/C; avoid random stock red/green.
  5. Derive interaction states (hover/active) by nudging L ±0.04–0.06, not opacity.
  6. Verify contrast for every text/bg pair you'll actually use.

Derive from the brief, not this file

The sample shows ramp mechanics, not a palette menu. Start from brand evidence, audience, ambient light, and roles; state why hue, neutral temperature, and accent energy fit. Missing evidence gets an approved assumption, not fashionable blue/violet/green.

Build a token system, not a swatch sheet

Keep three layers separate: primitives are named OKLCH stops; semantic roles name jobs such as text, surface, focus, and danger; component aliases reference one role, never a raw value. Define focus, non-color cues, and rendered pairs per status.

Non-negotiables

  • Build ramps in OKLCH with controlled lightness/chroma. Keep hue stable when useful; document hue correction when a constant ramp looks uneven or leaves gamut.
  • Give every hue a job. Do not add equally weighted colors for completeness or force a single-accent formula where evidence needs more.
  • Choose neutral temperature from content. Tinted and true-neutral ramps are valid when the role is explicit.
  • Verify WCAG on real pairs: body text ≥ 4.5:1, large text/UI/icons ≥ 3:1. Measure; WCAG 2.x ratios are this skill's AA thresholds, while legal applicability depends on jurisdiction. APCA is an additional reading check, not a substitute. CSS contrast-color() still ships only black/white; do not rely on it for brand pairs.
  • Don't convey meaning by color alone (color-blind users) — pair with icon/text.
  • Chroma tapers at lightness extremes. Max chroma is mid-L; near-white/black stops must drop C or look neon/muddy.

Measure the rendered pair, including gamut and state

Measure each rendered text, icon/UI, focus, and error pair in every theme/state; a passing body pair does not prove disabled or hover. For P3, retain hierarchy in the sRGB fallback and record gamut compression. If a hue cannot carry a legible role, use it as a mark/large accent and assign darker semantic ink.

Quick-start

Use assets/oklch-ramp.md for the generator/token skeleton. Keep primitives separate from roles; measure real text/UI pairs and emit hex fallbacks for legacy targets.

Reference files

  • references/oklch-and-ramps.md — OKLCH, ramp math, chroma taper, harmony, neutral tinting, P3/fallbacks. Read when generating a palette.
  • references/contrast-and-tokens.md — WCAG targets, contrast measurement, token architecture, interaction states, color-blind safety, light/dark sharing.
  • references/decision-records.md — meta-patterns + ADR rules for novel cases.

How to deliver

State hue/neutral strategy, semantic tokens, measured pairs, gamut fallbacks, and gaps; hand role mappings to theming.

Evidence

Record brief fit, primitive/role map, real contrast, non-color cues/gamut fallback, and theming handoff with evidence, reason, remediation.

Contract (generated)

Canonical detail: contract.json.

  • Route: A design needs a cohesive palette, color ramp, semantic tokens, or contrast repair (+1 in contract.json); avoid: The request is only to map existing semantic tokens between light, dark, or forced-colors themes (+1 in contract.json)
  • Exclude: Do not deliver an unmeasured palette (+1 in contract.json)
  • Stop / handoff: Pause when the required foreground-background roles are unknown (+1 in contract.json); receives [design-system-interview, deslop-ui, improve-existing-website, tasteroll] -> sends [theming, data-viz, a11y-pass]
  • Output: A brief-derived OKLCH ramp and semantic color-token system with measured pair evidence
  • Evidence: table_with_evidence with status, reason, remediation, evidence, provenance.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Color System AI skill do?

OKLCH color-system guidance for cohesive palettes and contrast-safe tokens. Use when choosing colors, generating ramps, fixing muddy palettes, creating semantic state colors, checking WCAG contrast, or defining theme variables.

Why use Color System on TypingMind?

Because you install it once and use it with any model. Color System is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Color System in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/KyaniteLabs/tastecheck/tree/main/skills/color-system. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Color System?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Color System?

As many as you like. As long as a model supports skills, you can use Color System with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Color System AI skill free?

Yes. It is published on GitHub by KyaniteLabs under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇