Taste Accessibility

Community

madebymustafa

taste-accessibility

Accessibility built into generation, not bolted on after: WCAG 2.2 AA compliance checks (contrast, keyboard, focus, labels, semantics, target size, reduced motion, autocomplete, error handling), baked into every page or component the agent builds. Run before calling any UI done. Stack on the design-taste core skill.

Overview

Publisher	madebymustafa
Repository	`design-taste`
Skill name	`taste-accessibility`
Stars	198
Forks	10
Bundled files	Instructions only
License	MIT
Links	Skill source Repository

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.
Self-contained
Everything the model needs lives in the instructions — no extra files to sync.
Open source
Published by madebymustafa on GitHub. Read the source before you install it.

Installation

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

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.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/madebymustafa/design-taste/tree/main/skills/taste-accessibility
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.

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/madebymustafa/design-taste.git /tmp/design-taste
mkdir -p .claude/skills
cp -r /tmp/design-taste/skills/taste-accessibility .claude/skills/taste-accessibility

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Taste Accessibility 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 Taste Accessibility 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 Taste Accessibility 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.

taste-accessibility

AI-generated UI is inaccessible by default. This skill makes accessible the default. Every rule below is applied during generation — not as a fix-up pass.

1. SEMANTIC STRUCTURE (The Code Skeleton)

One <h1> per page; heading levels never skip (h1 → h3 without h2 = failure); headings carry the page outline — test by reading only headings.
Landmarks: <header>, <nav>, <main>, <footer>, plus <aside>/<section> with accessible names where meaningful.
Lists for lists (<ul> for nav, features), <table> only for tabular data with <th scope> headers.
Buttons are <button>; links are <a href>; both have visible text. Icon-only controls get aria-label matching their visible tooltip.
Decorative images: alt="" (empty). Informative images: alt describing meaning, not appearance ("Chart: conversion fell from 18% to 9% in May" — not "a blue line chart").
aria-hidden="true" + tabindex="-1" on decorative SVG/decor; never aria-hidden on focusable elements.

2. KEYBOARD COMPLETE (Nothing Requires a Mouse or Touch)

Everything reachable: tab order follows visual order; no element focusable that isn't interactive (tabindex="0" only on purpose).
Focus visible everywhere: 2px ring, 3:1 contrast vs adjacent, outline-offset ≥ 2px. Never outline: none without a replacement.
Standard interactions: Enter activates buttons/links; Space toggles; Esc closes dialogs and menus.
Dialogs/popovers: focus moves into the dialog on open, traps within (or the platform's native pattern), returns to the trigger on close; overlay has role="dialog" + aria-modal="true" + labeled.
Skip link: first focusable element on every marketing page ("Skip to content"), visible on focus.
Custom components (menu, tabs, combobox, slider): implement ARIA patterns correctly (e.g., role="tablist" with aria-selected, arrow-key navigation) or use the platform element instead — a native <select> beats a broken custom one.

3. CONTRAST & TARGETS (Exact Numbers, WCAG 2.2)

Rule	Requirement
Normal text	≥ 4.5:1
Large text (≥24px / ≥18.66px bold)	≥ 3:1
UI components & focus	≥ 3:1 vs adjacent
Pointer target size	≥ 24×24px (AA 2.2), ≥ 44×44px recommended & for touch
Spacing exception	24px targets with ≥ 24px gaps to neighbors pass

Placeholder-only labels fail (4.1.2): every input needs a programmatically associated label (<label for> or aria-labelledby).
Error states: never color-only — text message + icon; errors announced (aria-live="assertive" region); error messages identify the field and give the fix, not "invalid".
Focus indicators on dark and light surfaces both; check focused state against ITS background, not page background.

4. FORMS (Generation Checklist)

Correct type per input (email, tel, password, number, search) + correct autocomplete tokens (email, name, credit-card-number, etc.).
required + aria-required; validation on :user-invalid styling (not just :invalid — don't punish pre-submit), with visible messages.
One column layout for anything ≥ 4 fields; labels above inputs; hint text associated (aria-describedby) not detached.
Multi-step wizards: progress announced, errors summarize at top, back/next keyboard-safe.

5. MOTION & COGNITION

Everything under prefers-reduced-motion (see taste-motion §7): no parallax, no carousel autoplay, no marquees.
No auto-playing media with sound; carousels pause on hover/focus; content doesn't move faster than the eye can read it.
Timeouts (sessions, auto-advance): ≥ 20s warning or user control (WCAG 2.2.1).
Plain language; expand abbreviations at first use; no emoji-only communication.

6. TESTING (Run These Checks Before Done)

Keyboard walk: Tab through the entire page — focus visible at every stop, order sane, nothing trapped.
Read with a screen reader (VoiceOver/NVDA): page title, heading outline, every control labeled, regions announced.
Contrast: sample every text/background pair used (at least: headings, body, muted, button text, placeholder, borders of inputs).
Resize: at 200% zoom (up to 1280px wide), no horizontal scroll, no text clipping.
Check aria-labels don't duplicate visible text unnecessarily; check no empty buttons/links exist.
Dark/light variants both pass (never assume one does).

7. CHECKLIST (Fail the Page If…)

Any heading level skipped.
Any input without an associated label.
Any interactive element unreachable by keyboard.
Any focus indicator invisible or outline: none without replacement.
Any text pair below 4.5:1 (3:1 for large) that isn't a deliberate, documented exception.
Any error communicated by color alone.
No prefers-reduced-motion handling.
Touch targets under 24px with less than 24px spacing.

Accessibility is not a feature. It is the interface working. Ship nothing that fails this checklist.

Frequently asked questions

What does the Taste Accessibility AI skill do?

Why use Taste Accessibility on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/madebymustafa/design-taste/tree/main/skills/taste-accessibility. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Taste Accessibility?

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 Taste Accessibility?

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

Is the Taste Accessibility AI skill free?

Yes. It is published on GitHub by madebymustafa 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

Mono Color

NousResearch

OrganizationPopular

Generate one- or two-ink editorial print poster images.

Pretext

Build creative browser demos with DOM-free text layout.

Sketch

Throwaway HTML mockups: 2-3 design variants to compare.

Tldraw Offline

Drive and script tldraw offline canvases with an agent.

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

taste-accessibility

1. SEMANTIC STRUCTURE (The Code Skeleton)

2. KEYBOARD COMPLETE (Nothing Requires a Mouse or Touch)

3. CONTRAST & TARGETS (Exact Numbers, WCAG 2.2)

4. FORMS (Generation Checklist)

5. MOTION & COGNITION

6. TESTING (Run These Checks Before Done)

7. CHECKLIST (Fail the Page If…)

Frequently asked questions

Related AI Skills

Set up your own AI workspace now