A skill your agent uses when asked to define a design token system, create tokens, document tokens, set up CSS custom properties, build a Tailwind token config, establish a spacing scale, define…

MITAuto-check: notesFrontend & Design

Install Form Tokens

skills CLI
$ npx skills add jeremylongshore/tons-of-skills-marketplace --skill form-tokens -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install jeremylongshore/tons-of-skills-marketplace form-tokens --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/jeremylongshore/tons-of-skills-marketplace.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/ai-agency/tonone/skills/form-tokens .claude/skills/form-tokens && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
form-tokens
GitHub stars
2.8k
Token cost
~6.9k tokens
SKILL.md length
1,303 words
Files
2
Skills in repo
3,342
Repo updated
First seen
Licence
MIT

At a glance

A skill your agent uses when asked to define a design token system, create tokens, document tokens, set up CSS custom properties, build a Tailwind token config, establish a spacing scale, define…

  • Works in 5 steps: Discovery → Token Architecture → Token Categories → …
  • Asked to define a design token system
  • SKILL.md covers Phase 1: Discovery, Phase 2: Token Architecture, Design Intelligence (via uiux) and Phase 3: Token Categories, plus 4 more sections
  • Calls python3

What it does

Form Tokens is an agent skill from jeremylongshore/tons-of-skills-marketplace. Use when asked to define a design token system, create tokens, document tokens, set up CSS custom properties, build a Tailwind token config, establish a spacing scale, define color semantics, or bridge design decisions to code. Examples: "set up design tokens", "define our token system", "create CSS variables for the design system", "document our color tokens", "establish a spacing scale".

Its SKILL.md is about 6.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `.claude-plugin/plugin.json`).

It sits in Frontend & Design, covering Design tokens, CSS and styling and Architecture decision records. It works with Tailwind CSS. The repository describes itself as: Model-agnostic agent-skills platform with a harness-free canonical layer, verified adapters, and the ccpi package manager. Explore at tonsofskills.com. The licence is MIT.

When your agent uses it

  • Asked to define a design token system
  • Document tokens
  • Set up CSS custom properties
  • Build a Tailwind token config

Example prompts

  • “set up design tokens”
  • “define our token system”
  • “create CSS variables for the design system”
  • “/form-tokens”

Requirements

  • Python 3
  • Pre-approved tools (allowed-tools): Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion

Workflow steps

5 steps, taken from the step headings in SKILL.md.

  1. Discovery
  2. Token Architecture
  3. Token Categories
  4. Output Format
  5. Dark Mode

What it can do on your machine

Read from SKILL.md and the folder at commit cfae287. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Read
    • Write
    • Edit
    • Bash
    • Glob
    • Grep
    • WebFetch
    • WebSearch
    • Task
    • TodoWrite

    …and 1 more on the same allowed-tools line.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • python3

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Form Tokens loads about 6.9k tokens when it runs. Until then it costs about 101 tokens; SKILL.md has 1,303 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~101
When it runs · the whole SKILL.md, loaded when a task matches
~6.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
    allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from jeremylongshore/tons-of-skills-marketplace at commit cfae287, republished under its MIT licence (© jeremylongshore). 1,303 words, ~6,880 tokens.

Download SKILL.mdSave it as .claude/skills/form-tokens/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
form-tokens
description
Use when asked to define a design token system, create tokens, document tokens, set up CSS custom properties, build a Tailwind token config, establish a spacing scale, define color semantics, or bridge design decisions to code. Examples: "set up design tokens", "define our token system", "create CSS variables for the design system", "document our color tokens", "establish a spacing scale".
allowed-tools
Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion
version
0.6.4
author
tonone-ai <hello@tonone.ai>
license
MIT

Form Tokens

You are Form — the visual designer on the Product Team.

Design tokens are the contract between design and code. They are not a deliverable — they are infrastructure. Every color, spacing value, and type size in the product should reference a token. This skill has 5 phases. Move through them in order. Do not skip phases.

Follow the output format defined in docs/output-kit.md — 40-line CLI max, box-drawing skeleton, unified severity indicators, compressed prose.


Phase 1: Discovery

Before producing any tokens, you need to understand what already exists and what constraints apply. Ask these questions. Group them naturally — do not fire them as a list.

Brand foundation
  • Has form-brand been run? Is there a brand brief with a defined palette, type system, and visual language?
  • If no brand brief exists, stop here. Run form-brand first. Tokens are downstream of brand decisions — defining tokens without a brand is building on sand.
Tech stack
  • What is the target stack? (CSS custom properties, Tailwind CSS, Style Dictionary, a design tool like Figma variables?)
  • Is there an existing token file or partial system to audit, or are we starting from zero?
  • Do tokens need to be exported to multiple formats (JSON, SCSS, Tailwind config, iOS Swift, Android XML)?
Platform targets
  • Which platforms need tokens? (Web, iOS, Android, email, print?)
  • Multi-platform targets require Style Dictionary or an equivalent build step — flag this early if relevant.
Existing constraints
  • Are there hardcoded hex values, magic numbers, or inline styles in the codebase right now? (These are the things tokens will replace.)
  • Is there a dark mode requirement today, or is it planned for the future? (The answer changes how semantic tokens are structured from day one.)

Done when: You know the brand state, the stack, the platforms, and whether dark mode is in scope.


Phase 2: Token Architecture

Before producing a single token, explain the two-tier model. Do not skip this explanation — it is why the system works, and teams who skip it break it later.

The two-tier model

Primitive tokens are raw values with no semantic meaning. They name a value, not a purpose.

css
--color-blue-100: #e6eeff;
--color-blue-200: #b3caff;
--color-blue-300: #80a8ff;
--color-blue-400: #4d85ff;
--color-blue-500: #0057ff;
--color-blue-600: #0047d6;
--color-blue-700: #0038ad;
--color-blue-800: #002884;
--color-blue-900: #001a5c;

--color-gray-50: #f9fafb;
--color-gray-100: #f3f4f6;
--color-gray-200: #e5e7eb;
--color-gray-300: #d1d5db;
--color-gray-400: #9ca3af;
--color-gray-500: #6b7280;
--color-gray-600: #4b5563;
--color-gray-700: #374151;
--color-gray-800: #1f2937;
--color-gray-900: #111827;

Primitives live in one place. They are never referenced directly in components.

Semantic tokens are meaning-bearing aliases that point to primitives. They name a purpose, not a value.

css
--color-action-primary: var(--color-blue-500);
--color-action-primary-hover: var(--color-blue-600);
--color-action-primary-active: var(--color-blue-700);
--color-text-default: var(--color-gray-900);
--color-text-muted: var(--color-gray-500);
--color-surface-default: var(--color-gray-50);

Semantic tokens are what components reference. Always.

The rule
Component → semantic token → primitive token → raw value

A button never references --color-blue-500 directly. It references --color-action-primary, which happens to point to --color-blue-500 today. When the brand color changes, one primitive changes, one semantic alias updates, and every component that references --color-action-primary updates automatically — with zero component changes.

This is the whole point of the system.

Confirm understanding before proceeding

Ask the team to confirm they understand the two-tier model and the rule above. If there is any hesitation, work through a concrete example from their existing UI before continuing.


Design Intelligence (via uiux)

After confirming token architecture (Phase 2), use the design system generator to seed initial token values:

bash
python3 -m form_agent.uiux design-system --product-type "{product_type}"

Use the generated design system output to:

  • Seed primitive color tokens from the industry-matched palette
  • Seed typography tokens from the recommended font pairing
  • Validate spacing and effect choices against the style recommendation

Phase 3: Token Categories

Define tokens in this order. Each category builds on the previous.


3.1 Color Primitives

Full palette scales for each hue in the brand. Use 50–900 steps. 500 is the base hue — not necessarily the one used most, but the reference point for the scale.

Derive values systematically from the brand's primary, secondary, neutral, and any accent colors. Do not invent hues that were not established in the brand brief.

css
/* ── Color Primitives ────────────────────────────────────── */

/* Brand primary */
--color-indigo-50: #eef2ff;
--color-indigo-100: #e0e7ff;
--color-indigo-200: #c7d2fe;
--color-indigo-300: #a5b4fc;
--color-indigo-400: #818cf8;
--color-indigo-500: #6366f1;
--color-indigo-600: #4f46e5;
--color-indigo-700: #4338ca;
--color-indigo-800: #3730a3;
--color-indigo-900: #312e81;

/* Neutral */
--color-gray-50: #f9fafb;
--color-gray-100: #f3f4f6;
--color-gray-200: #e5e7eb;
--color-gray-300: #d1d5db;
--color-gray-400: #9ca3af;
--color-gray-500: #6b7280;
--color-gray-600: #4b5563;
--color-gray-700: #374151;
--color-gray-800: #1f2937;
--color-gray-900: #111827;

/* Feedback palettes */
--color-green-50: #f0fdf4;
--color-green-500: #22c55e;
--color-green-700: #15803d;
--color-green-900: #14532d;

--color-yellow-50: #fefce8;
--color-yellow-500: #eab308;
--color-yellow-700: #a16207;
--color-yellow-900: #713f12;

--color-red-50: #fff1f2;
--color-red-500: #f43f5e;
--color-red-700: #be123c;
--color-red-900: #881337;

--color-blue-50: #eff6ff;
--color-blue-500: #3b82f6;
--color-blue-700: #1d4ed8;
--color-blue-900: #1e3a8a;

/* Absolute */
--color-white: #ffffff;
--color-black: #000000;

3.2 Color Semantic

Group semantic tokens by role. Name them by purpose, never by value.

Document use rules with every group: where each token appears, and where it must not.

css
/* ── Color Semantic: Surface ─────────────────────────────── */
/* Surfaces are backgrounds — page, panel, card, input, overlay */

--color-surface-page: var(--color-gray-50); /* Root page background */
--color-surface-default: var(--color-white); /* Cards, panels, modals */
--color-surface-raised: var(
  --color-white
); /* Elevated surfaces — use with shadow tokens */
--color-surface-sunken: var(
  --color-gray-100
); /* Input fields, well backgrounds */
--color-surface-overlay: var(--color-white); /* Dropdown menus, popovers */
--color-surface-subtle: var(
  --color-gray-100
); /* Hover states on non-interactive areas */

/* DO NOT use surface tokens for text, borders, or icons */

/* ── Color Semantic: Text ────────────────────────────────── */
/* Text tokens govern all typographic content */

--color-text-default: var(
  --color-gray-900
); /* Body copy, headings — default readable text */
--color-text-muted: var(
  --color-gray-500
); /* Secondary text, captions, placeholders */
--color-text-subtle: var(
  --color-gray-400
); /* Disabled text, decorative labels */
--color-text-inverse: var(--color-white); /* Text on dark/colored backgrounds */
--color-text-link: var(--color-indigo-600); /* Inline links */
--color-text-link-hover: var(--color-indigo-700); /* Inline links on hover */

/* DO NOT use --color-text-muted for interactive elements — use action tokens */

/* ── Color Semantic: Border ──────────────────────────────── */
/* Borders define structure — dividers, outlines, rings */

--color-border-default: var(
  --color-gray-200
); /* Default dividers, card edges */
--color-border-strong: var(
  --color-gray-300
); /* Emphasized dividers, section breaks */
--color-border-focus: var(
  --color-indigo-500
); /* Focus rings on interactive elements */
--color-border-error: var(--color-red-500); /* Error state borders on inputs */

/* DO NOT use --color-border-default for focus rings — always use --color-border-focus */

/* ── Color Semantic: Action ──────────────────────────────── */
/* Actions drive user intent — buttons, links, triggers */

--color-action-primary: var(--color-indigo-600);
--color-action-primary-hover: var(--color-indigo-700);
--color-action-primary-active: var(--color-indigo-800);
--color-action-primary-text: var(--color-white);

--color-action-secondary: var(--color-gray-100);
--color-action-secondary-hover: var(--color-gray-200);
--color-action-secondary-active: var(--color-gray-300);
--color-action-secondary-text: var(--color-gray-900);

--color-action-destructive: var(--color-red-600);
--color-action-destructive-hover: var(--color-red-700);
--color-action-destructive-text: var(--color-white);

--color-action-disabled-bg: var(--color-gray-100);
--color-action-disabled-text: var(--color-gray-400);

/* DO NOT hardcode hover/active states in component code — use action tokens */

/* ── Color Semantic: Feedback ────────────────────────────── */
/* Feedback communicates system status — alerts, banners, badges, toasts */

--color-feedback-success-bg: var(--color-green-50);
--color-feedback-success-border: var(--color-green-500);
--color-feedback-success-text: var(--color-green-700);
--color-feedback-success-icon: var(--color-green-500);

--color-feedback-warning-bg: var(--color-yellow-50);
--color-feedback-warning-border: var(--color-yellow-500);
--color-feedback-warning-text: var(--color-yellow-700);
--color-feedback-warning-icon: var(--color-yellow-500);

--color-feedback-error-bg: var(--color-red-50);
--color-feedback-error-border: var(--color-red-500);
--color-feedback-error-text: var(--color-red-700);
--color-feedback-error-icon: var(--color-red-500);

--color-feedback-info-bg: var(--color-blue-50);
--color-feedback-info-border: var(--color-blue-500);
--color-feedback-info-text: var(--color-blue-700);
--color-feedback-info-icon: var(--color-blue-500);

/* DO NOT use raw primitive colors in alert/toast components — always use feedback tokens */

3.3 Spacing Scale

All spacing derives from a single base unit: 8px. No arbitrary values.

css
/* ── Spacing ─────────────────────────────────────────────── */
/* Base unit: 8px. Every value is a multiple or half-multiple of 8. */

--space-1: 4px; /* 0.5× — tight internal padding, icon gaps */
--space-2: 8px; /* 1×  — default small gap, icon padding */
--space-3: 12px; /* 1.5× — compact list item padding */
--space-4: 16px; /* 2×  — default component padding */
--space-6: 24px; /* 3×  — section internal spacing */
--space-8: 32px; /* 4×  — card padding, form field gap */
--space-12: 48px; /* 6×  — section padding, large gap */
--space-16: 64px; /* 8×  — page section rhythm */
--space-24: 96px; /* 12× — hero spacing, major breakpoints */

Token names map to the multiplier (--space-4 = 4px × the multiplier naming convention tracks with Tailwind's default scale). Adjust naming to match your stack — what matters is the math, not the names.

Rule: if a spacing value is not on this scale, it does not go in the code. Raise a design question first.


3.4 Typography

Define font family, size scale, weight scale, and line-height scale as separate token groups.

css
/* ── Typography: Font Family ─────────────────────────────── */

--font-sans: "Inter", "Helvetica Neue", Arial, sans-serif;
--font-serif: "Merriweather", Georgia, serif; /* editorial use only */
--font-mono: "JetBrains Mono", "Fira Code", monospace; /* code, data */

/* ── Typography: Size Scale ──────────────────────────────── */
/* Modular scale — each step ×1.25 (major third) from 16px base */

--text-xs: 12px; /* Labels, captions, badges */
--text-sm: 14px; /* Secondary body, input fields */
--text-base: 16px; /* Primary body copy */
--text-lg: 18px; /* Large body, lead paragraphs */
--text-xl: 20px; /* Small headings */
--text-2xl: 24px; /* Section headings */
--text-3xl: 30px; /* Page headings */
--text-4xl: 36px; /* Hero subheadings */
--text-5xl: 48px; /* Hero headings */
--text-6xl: 60px; /* Display / marketing */

/* ── Typography: Weight Scale ────────────────────────────── */

--font-weight-regular: 400;
--font-weight-medium: 500;
--font-weight-semibold: 600;
--font-weight-bold: 700;

/* ── Typography: Line Height ─────────────────────────────── */

--leading-none: 1; /* Display type only — no space between lines */
--leading-tight: 1.25; /* Headings */
--leading-snug: 1.375; /* Subheadings, callouts */
--leading-normal: 1.5; /* Body copy — default */
--leading-relaxed: 1.625; /* Long-form reading */
--leading-loose: 2; /* Captions, labels with extra air */

/* ── Typography: Letter Spacing ──────────────────────────── */

--tracking-tight: -0.025em; /* Large headings */
--tracking-normal: 0; /* Default */
--tracking-wide: 0.025em; /* UI labels, buttons */
--tracking-wider: 0.05em; /* Caps labels, badges */
--tracking-widest: 0.1em; /* Overlines, legal text */

3.5 Border Radius
css
/* ── Border Radius ───────────────────────────────────────── */

--radius-none: 0; /* Sharp — tables, data-dense UIs */
--radius-sm: 2px; /* Subtle rounding — badges, tags */
--radius-base: 4px; /* Default — inputs, buttons */
--radius-md: 6px; /* Cards, panels */
--radius-lg: 8px; /* Modals, drawers */
--radius-xl: 12px; /* Feature cards, image containers */
--radius-2xl: 16px; /* Large surface rounding */
--radius-full: 9999px; /* Pill — tags, avatars, toggle thumbs */

3.6 Shadow / Elevation

Elevation communicates layer height. Use consistently — don't reach for a deeper shadow for emphasis when border or color is the right tool.

css
/* ── Shadow / Elevation ──────────────────────────────────── */
/* Level 0: flat — no shadow. Used for page and panel backgrounds. */

--shadow-sm: 0 1px 2px 0 rgb(0 0 0 / 0.05);
/* Level 1 — Buttons, inputs, subtle surface lift */

--shadow-base: 0 1px 3px 0 rgb(0 0 0 / 0.1), 0 1px 2px -1px rgb(0 0 0 / 0.1);
/* Level 2 — Cards, dropdowns, date pickers */

--shadow-md: 0 4px 6px -1px rgb(0 0 0 / 0.1), 0 2px 4px -2px rgb(0 0 0 / 0.1);
/* Level 3 — Floating panels, contextual menus */

--shadow-lg: 0 10px 15px -3px rgb(0 0 0 / 0.1), 0 4px 6px -4px rgb(0 0 0 / 0.1);
/* Level 4 — Modals, dialogs */

--shadow-xl:
  0 20px 25px -5px rgb(0 0 0 / 0.1), 0 8px 10px -6px rgb(0 0 0 / 0.1);
/* Level 5 — Full-screen overlays, toast notifications */

3.7 Z-Index (Named Layers)

Z-index wars happen when values are chosen arbitrarily. Named layers end them.

css
/* ── Z-Index ─────────────────────────────────────────────── */

--z-base: 0; /* Normal document flow */
--z-raised: 10; /* Sticky headers, floating action buttons */
--z-dropdown: 100; /* Dropdown menus, autocomplete lists */
--z-sticky: 200; /* Sticky navigation */
--z-overlay: 300; /* Backdrop overlays (behind modals) */
--z-modal: 400; /* Modals, dialogs, drawers */
--z-toast: 500; /* Toast notifications (above modals) */
--z-tooltip: 600; /* Tooltips (above everything interactive) */

Rule: never use a raw integer for z-index in component code. Always reference a named layer. If the existing layers do not fit the need, add a named layer here — do not hardcode 9999.


Show full SKILL.md (519 more words)Show less
3.8 Motion
css
/* ── Motion: Duration ────────────────────────────────────── */

--duration-instant: 0ms; /* Immediate — no animation (reduced-motion fallback) */
--duration-fast: 100ms; /* Micro-interactions — button press, toggle */
--duration-normal: 200ms; /* Standard — state transitions, hover */
--duration-slow: 300ms; /* Deliberate — panel open, modal enter */
--duration-slower: 500ms; /* Narrative — page transitions, hero animations */

/* ── Motion: Easing ──────────────────────────────────────── */

--ease-linear: linear;
--ease-in: cubic-bezier(0.4, 0, 1, 1); /* Exits — elements leaving the screen */
--ease-out: cubic-bezier(0, 0, 0.2, 1); /* Entrances — elements arriving */
--ease-in-out: cubic-bezier(
  0.4,
  0,
  0.2,
  1
); /* State changes — two-way transitions */
--ease-spring: cubic-bezier(
  0.34,
  1.56,
  0.64,
  1
); /* Playful spring — toggles, thumbs */

Use --ease-out for entrances (user is waiting for something to appear — make it feel immediate). Use --ease-in for exits (user triggered the dismissal — don't slow them down). Use --ease-in-out for transitions where direction is ambiguous.

Always implement @media (prefers-reduced-motion: reduce) — swap durations to --duration-instant and easing to --ease-linear.


Phase 4: Output Format

CSS custom properties block

Deliver all tokens in a single :root block (light mode defaults), organized by category with clear section comments. The order from Phase 3 is the output order.

css
:root {
  /* Color Primitives */
  /* ... */

  /* Color Semantic: Surface */
  /* ... */

  /* Color Semantic: Text */
  /* ... */

  /* Color Semantic: Border */
  /* ... */

  /* Color Semantic: Action */
  /* ... */

  /* Color Semantic: Feedback */
  /* ... */

  /* Spacing */
  /* ... */

  /* Typography */
  /* ... */

  /* Border Radius */
  /* ... */

  /* Shadow */
  /* ... */

  /* Z-Index */
  /* ... */

  /* Motion */
  /* ... */
}
Tailwind config mapping (if applicable)

If the project uses Tailwind CSS, produce a tailwind.config.js that maps tokens to Tailwind's theme.extend using var() references. Do not duplicate raw values — Tailwind reads from the custom properties defined in :root.

js
// tailwind.config.js
module.exports = {
  theme: {
    extend: {
      colors: {
        "action-primary": "var(--color-action-primary)",
        "action-primary-hover": "var(--color-action-primary-hover)",
        "action-secondary": "var(--color-action-secondary)",
        "text-default": "var(--color-text-default)",
        "text-muted": "var(--color-text-muted)",
        "surface-page": "var(--color-surface-page)",
        "surface-default": "var(--color-surface-default)",
        "border-default": "var(--color-border-default)",
        "border-focus": "var(--color-border-focus)",
        "feedback-success-bg": "var(--color-feedback-success-bg)",
        "feedback-error-bg": "var(--color-feedback-error-bg)",
        // ... complete the mapping
      },
      spacing: {
        1: "var(--space-1)",
        2: "var(--space-2)",
        3: "var(--space-3)",
        4: "var(--space-4)",
        6: "var(--space-6)",
        8: "var(--space-8)",
        12: "var(--space-12)",
        16: "var(--space-16)",
        24: "var(--space-24)",
      },
      fontFamily: {
        sans: "var(--font-sans)",
        serif: "var(--font-serif)",
        mono: "var(--font-mono)",
      },
      fontSize: {
        xs: "var(--text-xs)",
        sm: "var(--text-sm)",
        base: "var(--text-base)",
        lg: "var(--text-lg)",
        xl: "var(--text-xl)",
        "2xl": "var(--text-2xl)",
        "3xl": "var(--text-3xl)",
        "4xl": "var(--text-4xl)",
        "5xl": "var(--text-5xl)",
        "6xl": "var(--text-6xl)",
      },
      borderRadius: {
        none: "var(--radius-none)",
        sm: "var(--radius-sm)",
        DEFAULT: "var(--radius-base)",
        md: "var(--radius-md)",
        lg: "var(--radius-lg)",
        xl: "var(--radius-xl)",
        "2xl": "var(--radius-2xl)",
        full: "var(--radius-full)",
      },
      boxShadow: {
        sm: "var(--shadow-sm)",
        DEFAULT: "var(--shadow-base)",
        md: "var(--shadow-md)",
        lg: "var(--shadow-lg)",
        xl: "var(--shadow-xl)",
      },
      zIndex: {
        base: "var(--z-base)",
        raised: "var(--z-raised)",
        dropdown: "var(--z-dropdown)",
        sticky: "var(--z-sticky)",
        overlay: "var(--z-overlay)",
        modal: "var(--z-modal)",
        toast: "var(--z-toast)",
        tooltip: "var(--z-tooltip)",
      },
      transitionDuration: {
        instant: "var(--duration-instant)",
        fast: "var(--duration-fast)",
        normal: "var(--duration-normal)",
        slow: "var(--duration-slow)",
        slower: "var(--duration-slower)",
      },
    },
  },
};
Style Dictionary (if multi-platform)

If tokens need to reach iOS or Android, recommend Style Dictionary as the build step. Produce the source JSON token file and a config.json. The CSS custom properties block above becomes one of the output formats — not the source of truth. Style Dictionary JSON is the source.


Phase 5: Dark Mode

Dark mode is not light mode inverted. Every semantic token needs an explicit value for dark mode. Explain this before proceeding.

The wrong approach
css
/* WRONG — DO NOT DO THIS */
html[data-theme="dark"] {
  filter: invert(1) hue-rotate(180deg);
}

This inverts images, photos, and icons. It produces incorrect colors for feedback states. It is a hack, not a system.

The right approach

Redefine every semantic token under a [data-theme="dark"] selector (or @media (prefers-color-scheme: dark) — choose one strategy and stay consistent). Primitives do not change. Semantic tokens point to different primitives.

css
[data-theme="dark"] {
  /* Surface */
  --color-surface-page: var(--color-gray-900);
  --color-surface-default: var(--color-gray-800);
  --color-surface-raised: var(--color-gray-800);
  --color-surface-sunken: var(--color-gray-900);
  --color-surface-overlay: var(--color-gray-800);
  --color-surface-subtle: var(--color-gray-700);

  /* Text */
  --color-text-default: var(--color-gray-50);
  --color-text-muted: var(--color-gray-400);
  --color-text-subtle: var(--color-gray-600);
  --color-text-inverse: var(--color-gray-900);
  --color-text-link: var(--color-indigo-400);
  --color-text-link-hover: var(--color-indigo-300);

  /* Border */
  --color-border-default: var(--color-gray-700);
  --color-border-strong: var(--color-gray-600);
  --color-border-focus: var(--color-indigo-400);
  --color-border-error: var(--color-red-400);

  /* Action */
  --color-action-primary: var(--color-indigo-500);
  --color-action-primary-hover: var(--color-indigo-400);
  --color-action-primary-active: var(--color-indigo-300);
  --color-action-primary-text: var(--color-white);

  --color-action-secondary: var(--color-gray-700);
  --color-action-secondary-hover: var(--color-gray-600);
  --color-action-secondary-active: var(--color-gray-500);
  --color-action-secondary-text: var(--color-gray-100);

  --color-action-destructive: var(--color-red-500);
  --color-action-destructive-hover: var(--color-red-400);
  --color-action-destructive-text: var(--color-white);

  --color-action-disabled-bg: var(--color-gray-800);
  --color-action-disabled-text: var(--color-gray-600);

  /* Feedback */
  --color-feedback-success-bg: var(--color-green-900);
  --color-feedback-success-border: var(--color-green-500);
  --color-feedback-success-text: var(--color-green-50);
  --color-feedback-success-icon: var(--color-green-500);

  --color-feedback-warning-bg: var(--color-yellow-900);
  --color-feedback-warning-border: var(--color-yellow-500);
  --color-feedback-warning-text: var(--color-yellow-50);
  --color-feedback-warning-icon: var(--color-yellow-500);

  --color-feedback-error-bg: var(--color-red-900);
  --color-feedback-error-border: var(--color-red-500);
  --color-feedback-error-text: var(--color-red-50);
  --color-feedback-error-icon: var(--color-red-500);

  --color-feedback-info-bg: var(--color-blue-900);
  --color-feedback-info-border: var(--color-blue-500);
  --color-feedback-info-text: var(--color-blue-50);
  --color-feedback-info-icon: var(--color-blue-500);
}

Notice: dark mode feedback tokens do NOT use -50 background primitives (which are near-white). They use -900 primitives (deep, desaturated). Dark mode is not a mirror — it is a separate set of decisions.

Dark mode checklist
[ ] Every semantic token has an explicit dark value
[ ] Feedback backgrounds use deep palette values, not inverted light values
[ ] Text muted color passes 3:1 contrast on dark surface
[ ] Primary action color passes 4.5:1 contrast on dark surface
[ ] Focus ring is visible on all dark surfaces
[ ] No filter: invert() anywhere in the codebase
[ ] prefers-color-scheme and data-theme strategies are consistent — pick one

Anti-Patterns

  • Semantic tokens that describe values, not purpose — --color-blue instead of --color-action-primary. When the brand color changes from blue to teal, --color-blue breaks. --color-action-primary does not.
  • Components that hardcode hex values — any #hex in a component file is a token debt. Every hex in a component is a future theming bug.
  • Dark mode as filter: invert() — inverts photos, corrupts feedback colors, and signals the team has no token system.
  • Dark mode as "just swap light and dark" — naive inversion produces incorrect contrast ratios and broken feedback states. Dark mode requires explicit decisions.
  • Spacing values that aren't on the scale — margin: 13px is a design decision that never got made. Surface it, snap it to the scale.
  • No z-index system — z-index: 9999 is a symptom. Named layers are the cure.
  • Semantic tokens that skip the primitive tier — --color-action-primary: #4F46E5 works, but it breaks the one-edit-propagates-everywhere guarantee. Always alias through a primitive.
  • Incomplete dark token sets — a semantic token without a dark value defaults to its light value in dark mode. This is always wrong, rarely intentional, and hard to catch in review.

Delivery Checklist

Before handing the token system to engineering, verify:

[ ] Brand brief confirmed — tokens derive from documented brand decisions
[ ] Two-tier model in place — primitives and semantic layers both exist
[ ] No component in the codebase references a primitive token directly
[ ] All 8 token categories defined: color primitives, color semantic, spacing,
    typography, border radius, shadow, z-index, motion
[ ] Spacing scale derives from 8px base — no arbitrary values
[ ] Every semantic token has documented use rules (where it appears, where it must not)
[ ] Dark mode: explicit values for every semantic token
[ ] Dark mode: feedback tokens use deep palette values, not inverted light values
[ ] Tailwind config (or equivalent) produced if applicable
[ ] Reduced-motion handled in motion tokens
[ ] Token file delivered as a single importable CSS file (or Style Dictionary source if multi-platform)

Delivery

If output exceeds the 40-line CLI budget, invoke /atlas-report with the full findings. The HTML report is the output. CLI is the receipt — box header, one-line verdict, top 3 findings, and the report path. Never dump analysis to CLI.

© jeremylongshore, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in plugins/ai-agency/tonone/skills/form-tokens of jeremylongshore/tons-of-skills-marketplace.

  • SKILL.md
  • .claude-plugin/plugin.json

Open the folder on GitHubat commit cfae287

Compare with similar skills

Form Tokens next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Form Tokens compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Form Tokens this skilljeremylongshore/tons-of-skills-marketplace2.8k—~6.9kAutomated safety check: NotesMIT
MCP Developmentcoollabsio/coolify63k1 repos~949Automated safety check: PassMIT
Tailwind Design SystemSuFxGIT/scoutarr1155 repos~5.6kAutomated safety check: PassNone
Extract DesignManavarya09/design-extract4.2k—~786Automated safety check: NotesMIT
Tailwind Theme Builderjezweb/claude-skills1.1k1 repos~3.2kAutomated safety check: PassMIT
Tailwind Design Systemjh941213/my-cc-harness12510 repos~4.7kAutomated safety check: PassNone

Similar skills

  • MCP Development

    coollabsio/coolify

    A skill your agent uses for Laravel MCP development. An agent skill from coollabsio/coolify.

    63k GitHub starsUsed in 1 repo~949 tokens
    Frontend & DesignAuto-check passed
  • Tailwind Design System

    SuFxGIT/scoutarr

    Build scalable design systems with Tailwind CSS v4, design tokens, component libraries, and responsive patterns.

    115 GitHub starsUsed in 5 repos~5.6k tokens
    Frontend & DesignAuto-check passed
  • Extract Design

    Manavarya09/design-extract

    Extract the full design language from any website URL. An agent skill from Manavarya09/design-extract.

    4.2k GitHub stars~786 tokensUpdated 2 days ago
    Frontend & DesignAuto-check: notes
  • Tailwind Theme Builder

    jezweb/claude-skills

    Set up Tailwind v4 + shadcn/ui themed UI with dark mode. An agent skill from jezweb/claude-skills.

    1.1k GitHub starsUsed in 1 repo~3.2k tokens
    Frontend & DesignAuto-check passed
  • Tailwind Design System

    jh941213/my-cc-harness

    Build scalable design systems with Tailwind CSS, design tokens, component libraries, and responsive patterns.

    125 GitHub starsUsed in 10 repos~4.7k tokens
    Frontend & DesignAuto-check passed
  • Preline Theme Generator

    htmlstreamofficial/preline

    Generates light and dark Preline theme CSS from a brand color or mood, previews it, and validates token coverage through bundled local scripts.

    6.5k GitHub stars~619 tokensUpdated 1 mo ago
    Frontend & DesignAuto-check: warnings

More from jeremylongshore/tons-of-skills-marketplace

All 3,342 skills in this repo
  • Performing Security Code Review

    jeremylongshore/tons-of-skills-marketplace

    Execute this skill enables AI assistant to conduct a security-focused code review using the security-agent plugin.

    2.8k GitHub starsUsed in 2 repos~1.3k tokens
    Auto-check: notes
  • Adapting Transfer Learning Models

    jeremylongshore/tons-of-skills-marketplace

    Build this skill automates the adaptation of pre-trained machine learning models using transfer learning techniques.

    2.8k GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Agent Context Loader

    jeremylongshore/tons-of-skills-marketplace

    Execute proactive auto-loading: automatically detects and loads agents.md files.

    2.8k GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Aggregating Performance Metrics

    jeremylongshore/tons-of-skills-marketplace

    Aggregate and centralize performance metrics from applications, systems, databases, caches, and services.

    2.8k GitHub stars~1.2k tokensUpdated today
    Auto-check passed
  • Analyzing Capacity Planning

    jeremylongshore/tons-of-skills-marketplace

    Execute this skill enables AI assistant to analyze capacity requirements and plan for future growth.

    2.8k GitHub stars~947 tokensUpdated today
    Auto-check passed
  • Analyzing Database Indexes

    jeremylongshore/tons-of-skills-marketplace

    Process use when you need to work with database indexing. An agent skill from jeremylongshore/tons-of-skills-marketplace.

    2.8k GitHub stars~2k tokensUpdated today
    Auto-check passed

Works with

Questions about Form Tokens

What does Form Tokens do?

A skill your agent uses when asked to define a design token system, create tokens, document tokens, set up CSS custom properties, build a Tailwind token config, establish a spacing scale, define…. Form Tokens is an agent skill from jeremylongshore/tons-of-skills-marketplace. Use when asked to define a design token system, create tokens, document tokens, set up CSS custom properties, build a Tailwind token config, establish a spacing scale, define color semantics, or bridge design decisions to code.

When should I use Form Tokens?

Form Tokens fits situations like: asked to define a design token system; document tokens; set up CSS custom properties; build a Tailwind token config.

How do I install Form Tokens in Claude Code?

Run `npx skills add jeremylongshore/tons-of-skills-marketplace --skill form-tokens -a claude-code`. Or copy the skill folder (plugins/ai-agency/tonone/skills/form-tokens in jeremylongshore/tons-of-skills-marketplace) into .claude/skills/form-tokens in your project. Claude Code loads it when a task matches its description.

How do I install Form Tokens in Codex?

Run `npx skills add jeremylongshore/tons-of-skills-marketplace --skill form-tokens -a codex`. Or copy the skill folder (plugins/ai-agency/tonone/skills/form-tokens in jeremylongshore/tons-of-skills-marketplace) into .agents/skills/form-tokens in your project. Codex loads it when a task matches its description.

Can I use Form Tokens in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add jeremylongshore/tons-of-skills-marketplace --skill form-tokens -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/form-tokens, .gemini/skills/form-tokens, .github/skills/form-tokens and .opencode/skills/form-tokens in your project.

What does Form Tokens need to run?

Going by SKILL.md and its folder, Form Tokens needs the command-line tools its instructions call (python3). Our summary lists: Python 3. Its frontmatter pre-approves these tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion.

Does Form Tokens access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Form Tokens safe to install?

Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Form Tokens use?

Form Tokens is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Form Tokens use?

About 6.9k tokens (SKILL.md is roughly 28k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Form Tokens?

Skills that share tags, products or a category with Form Tokens: MCP Development (coollabsio/coolify, 63k stars), Tailwind Design System (SuFxGIT/scoutarr, 115 stars), Extract Design (Manavarya09/design-extract, 4.2k stars) and Tailwind Theme Builder (jezweb/claude-skills, 1.1k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Form Tokens?

jeremylongshore (a GitHub user) maintains it in jeremylongshore/tons-of-skills-marketplace, which has 2,827 GitHub stars. The repository holds 3,342 skills in this directory. The repository was last updated on October 10, 2026.

Source: jeremylongshore/tons-of-skills-marketplace on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.