Theming logo

Theming

Organization
KyaniteLabs
theming

Use when semantic roles must map coherently across light, dark, forced-colors, or saved theme preferences without contrast or no-flash regressions.

Overview

PublisherKyaniteLabs
Repositorytastecheck
Skill nametheming
Stars
10
Forks
7
Bundled files
7
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.

  • 7 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 Theming 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/theming
  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/theming .claude/skills/theming
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Theming 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 Theming 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 Theming 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.

Theming

A theme system keeps component roles stable while modes remap values. Components reference semantic tokens such as --color-bg, --color-text, and --color-focus; measure every mode before claiming accessibility. Dark is not inverted light; authored high contrast is not forced colors.

The decision order

  1. Semantic tokens first. Define roles (--color-bg, --color-surface-1/2 (-3 if elevation needs it), --color-text, --color-text-muted, --color-border, --color-primary, --color-primary-hover, --color-primary-ink, --color-accent, --color-accent-ink, --color-success, --color-error, --color-warning, --color-info, --color-focus — canonical contract names). Components use these only; palette comes from color-system.
  2. Choose the product baseline. Start with the default supported mode; set structure and role relationships there.
  3. Tune every additional mapping, never invert it. For dark, use a lighter-per-level surface ramp, comfortable text contrast, and retuned accents. See references/surfaces-and-elevation.md and color-and-contrast.md.
  4. Separate authored contrast from forced colors. prefers-contrast: more may select a higher-contrast mapping; forced-colors is system-controlled, so honor its colors.
  5. Resolve preferences deliberately. Use prefers-color-scheme, persist selectable themes, and set color-scheme so native controls follow.
  6. Verify each theme on its actual surfaces.

Vary mappings, preserve contract

Starter values show role mapping, not a dark recipe. Derive mappings from roles, ambient conditions, brand, and platform constraints; keep role names, forced-colors respect, preference precedence, and measured pairs intact. Missing roles require named recovery; never invent component colors.

Non-negotiables

  • One semantic-token source; components never hard-code colors. Re-theming remaps tokens only.
  • Choose light surfaces from context. Pure or softened endpoints can both be valid; test contrast, glare, reading comfort, and preferences.
  • Dark is tuned, not inverted: near-black is reliable, but pure black can suit OLED, projection, or a stark system. Distinguish elevation with surfaces/borders, tune text below glare-inducing white where appropriate, and remeasure accents.
  • Ship a default and resolution policy. A supported user choice persists and outranks OS preference; a single-theme product needs no toggle.
  • color-scheme declared per theme (native controls/scrollbars follow).
  • Respect forced-colors/high-contrast — don't override system colors; use forced-color-adjust deliberately and keep focus visible.
  • Contrast re-verified per theme (body ≥4.5, large/UI ≥3) against that theme's surfaces — a color that passes in light often fails in dark, and vice-versa.

Resolution and recovery

Separate role names, theme values, and selection. Document precedence: system forced-colors; supported user choice; OS preference; application default. For a control, define storage and fallback.

Forced-colors is an environment, not a palette: keep system colors/native controls authoritative and preserve warnings, errors, selections, and focus with text, icons, state, or shape. If a role is absent, record a failed recovery row and stop; reusing an accent changes its meaning.

Apply the theme early to minimize wrong-theme flash. State pre-paint behavior, unavailable saved choices, and native-control color-scheme. No-flash needs a cold-load observation naming theme, artifact, and timing.

Quick-start

Use assets/theme-starter.css; preserve roles, tuned mappings, forced-colors cooperation, and early saved-preference resolution.

Early saved-theme resolution: when a persistent control exists, apply the validated override before render. In a CSP-compatible implementation, this is commonly a small nonced or hashed <head> script rather than a deferred bundle:

html
<script>/* inline, in <head>, before CSS paint */
  try {
    const t = localStorage.getItem("theme");
    if (t === "dark" || t === "light") document.documentElement.dataset.theme = t;
  } catch {
    /* Storage is unavailable; CSS keeps the OS/application default authoritative. */
  }
</script>

Browser chrome, selection, and custom scrollbars are optional polish, not completion gates.

light-dark() shorthand

When color-scheme drives the switch (no persisted override), light-dark() avoids media-query duplication: background: light-dark(oklch(96% .02 78), oklch(18% .024 64)). Broad browser support (2024+); pair with [data-theme] so explicit choice and forced-colors override. It is not the sole mechanism with a manual toggle.

Reference files

  • references/surfaces-and-elevation.md — dark ramp, elevation, borders, shadow limits.
  • references/color-and-contrast.md — dark accents, state colors, contrast, OKLCH.
  • references/light-and-contrast.md — light mappings, relationships, accents, rendered checks.
  • references/high-contrast.md — prefers-contrast + forced-colors; override limits.
  • references/decision-records.md — meta-patterns + ADR rules.

Ledger

Close with five-field ledger. Start Reason with the check ID; name mode/environment in Evidence.

StatusReasonRemediationEvidenceProvenance
role-map — semantic roles across supported themes
supported-maps — each requested mode is tuned rather than inverted
resolution — preference, no-flash, and native-control behavior
forced-colors — high-contrast environmental behavior
contrast-handoff — per-theme proof and accessibility boundary

How to deliver

Report roles, mappings, resolution, wrong-theme-flash behavior, and per-theme contrast, with scoped-environment and forced-colors proof when applicable. Hand source ramps, type, technical a11y, and cognitive concerns to adjacent skills. Provenance is credited and independent.

Contract (generated)

Canonical detail: contract.json.

  • Route: An interface needs coherent light, dark, high-contrast, or forced-colors mappings (+1 in contract.json); avoid: The request is to invent the source palette rather than map semantic roles (+1 in contract.json)
  • Exclude: Do not invert a light palette and call it dark mode (+1 in contract.json)
  • Stop / handoff: Pause when semantic roles are missing and theme values would be raw-color copies (+1 in contract.json); receives [color-system, design-system-interview, improve-existing-website, tasteroll] -> sends [a11y-pass, cognitive-a11y, responsive-layout, component-states]
  • Output: Semantic mappings for every supported theme, plus forced-colors behavior, preference policy, and measured pairs
  • 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 Theming AI skill do?

Use when semantic roles must map coherently across light, dark, forced-colors, or saved theme preferences without contrast or no-flash regressions.

Why use Theming on TypingMind?

Because you install it once and use it with any model. Theming 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 Theming in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/KyaniteLabs/tastecheck/tree/main/skills/theming. 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 Theming?

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 Theming?

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

Is the Theming 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 👇