Cds Code logo

Cds Code

Organization
coinbase
cds-code

Provides a workflow for building and reviewing high-quality React and React Native UI with the Coinbase Design System (CDS). Use this skill to create UI or make visual changes to existing UI with CDS components, tokens, style props, icons, and illustrations. Use it for explicit CDS-adherence audits and reviews, not non-visual logic, data, analytics, navigation, tests, generated code, or refactoring. Examples: "build this screen", "fix spacing on this component", "replace this UI with CDS components", "perform a CDS audit on our changes", "check our codebase for CDS adherence", "does this feature use CDS well?"

Overview

Publishercoinbase
Repositorycds
Skill namecds-code
Stars
505
Forks
107
Bundled files
8
LicenseApache-2.0
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.

  • 8 bundled files

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

  • Open source

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

Installation

Install the Cds Code 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.
  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/coinbase/cds.git /tmp/cds
mkdir -p .claude/skills
cp -r /tmp/cds/skills/cds-code .claude/skills/cds-code
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Cds Code 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 Cds Code 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 Cds Code 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.

CDS Code Skill

On every request

Before responding, determine what the user needs:

Coding — the user wants to create new UI or make visual changes to existing UI with CDS → follow the Coding Workflow.

Review — the user explicitly asks to audit, review, or check existing code for CDS adherence → read guidelines/code-review.md and follow it instead.

Out of scope — the user wants non-visual work such as business logic, data handling, analytics, navigation wiring, tests, generated code, or refactoring → do not follow this skill's coding or review workflows.

Default in-scope work to coding. Only treat the user's request as a review or audit if the intent is explicit. Building or updating UI is the primary use case for this skill.

Initialization

Run this once per session, before doing anything else.

Run the discovery script: scripts/discover-cds-packages.sh

Its output tells you:

  • The CDS Runtime (web or mobile) — use this value as the platform argument for the CDS MCP server if it is needed.
  • Every installed CDS package: its name, version, and valid export subpaths — these import paths are the ONLY ALLOWED PATHS when importing from CDS packages

If the script cannot be run, much of the information it provides can be determined via manual inspection:

  • Infer the platform by inspecting existing CDS imports in the project's source code
  • Find valid import paths by reading the exports field of the package.json of installed CDS packages in node_modules

Coding Workflow

For in-scope coding tasks, follow these steps in order.

YOU MUST perform steps 1 and 2 before writing any code!

Step 1: Prepare CDS documentation

For any CDS documentation needs, use either of the following tools. If neither are available, let the user know but continue — documentation is helpful but not required.

  • Activate the cds-docs skill OR...
  • If the cds-docs skill is not configured, try calling the CDS MCP server list-cds-routes tool.

Then read the platform-specific docs (using the runtime detected in Initialization):

  • getting-started/styling
  • getting-started/theming

Step 2: Identify the appropriate components

Use guidelines/components.md to help identify the appropriate CDS components for the task. The guidelines file will cover most use cases, but you may optionally browse the CDS docs for the full list of supported CDS components.

If you decide your task will require icons (Icon or IconButton) or illustrations (SpotSquare, Pictogram, HeroSquare, etc.) please read the corresponding guidelines files for more details.

IconsIllustrations
guidelines/icons.mdguidelines/illustrations.md

If no CDS component fits your use case, you may fall back to the following options in this order of priority:

  1. search for a relevant and reusable React component from the project's codebase to use
  2. build your own custom React component using CDS primitives as building blocks
  3. use the native platform's JSX elements (div, View, etc.) for bespoke UI as a last resort

IMPORTANT: Always inform the user which CDS components you are planning to use before moving on to Step 3.

Step 3: Optionally read component docs

For any CDS component you plan to use, retrieve and read their documentation (see Step 1 in this workflow for more details on docs setup).

If documentation is not retrievable for any reason, the published type definitions for the component may be used to determine the full props API a component affords. This is no substitute for reading the documentation, but it can be a useful fallback when documentation is not available.

Step 4: Execute the task (writing code)

Now create or update the UI with proper CDS components and usage.

Package scope

The package name may vary between projects. Different repos may install CDS under different scopes. Always match the full CDS package name(s) as determined in the initialization step. If the project already has CDS imports in existing code, match whatever scope those files use.

Using the Design System

In most cases, you should avoid using inline style objects or CSS classNames (web only). Through these methods it is very easy to make common mistakes like using hardcoded property values instead of the CDS design tokens. Doing so would break the component's connection to the CDS theme.

If you must use a style object or a CSS className, you can still access the CDS theme either through the useTheme hook or by CSS variables (web only).

Most CDS components implement an API that conveniently allows you to apply CDS design tokens, internally we call these 'style props'.

In cds-web, style props essentially act as an API for applying atomic CSS classes, much like Tailwind's utility classes which are so prevelant in the web ecosystem.

You should prefer setting these style props for styling components over setting custom style via inline styles or CSS.

Why this matters: When you set font, color, textAlign, or other typography properties through style instead of props, the component loses its connection to the CDS theme. For example, setting fontSize and fontWeight via style without a font prop means the CDS font family never applies -- the text falls back to inherit and may render in the wrong typeface.

You should check a component's documentation which includes a props table to verify the available API.

Examples of opportunities to use style props over inline styles:

Instead of styleUse the prop
style={{ color: "var(--color-fgMuted)" }}color="fgMuted"
style={{ fontSize: 12, fontWeight: 500, lineHeight: "16px" }}font="caption" (or the matching CDS font token)
style={{ textAlign: "center" }}textAlign="center"
style={{ textTransform: "uppercase" }}textTransform="uppercase"
style={{ display: "flex", flexDirection: "column" }}Use VStack, or flexDirection="column" on Box
style={{ gap: 8 }}gap={1}
style={{ padding: 16 }}padding={2}
style={{ backgroundColor: "..." }}background="bgAlternate" (or semantic token)

Step 5: Validate changes

Your task will be complete if:

  1. You performed skill initialization and explicitly identified the specific CDS components you would use
  2. Your changes DO NOT include any raw rgb/hex/etc color values
  3. Your changes DO NOT use any raw pixel values for spacing properties (padding, margin, gap, border radius). Explicit layout dimensions like width or height set to specific designer-specified values are acceptable.
  4. Your changes DO NOT import any depreacted CDS components or hooks.
  5. Your changes use components' style props (e.g. font, color, background, textTransform, paddingX, gap) instead of customization via inline style objects or with CSS classNames.
  6. All import paths are valid CDS package exports, determined in initialization
  7. The project's linting/typechecking/formatting tasks are passing

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 Cds Code AI skill do?

Provides a workflow for building and reviewing high-quality React and React Native UI with the Coinbase Design System (CDS). Use this skill to create UI or make visual changes to existing UI with CDS components, tokens, style props, icons, and illustrations. Use it for explicit CDS-adherence audits and reviews, not non-visual logic, data, analytics, navigation, tests, generated code, or refactoring. Examples: "build this screen", "fix spacing on this component", "replace this UI with CDS components", "perform a CDS audit on our changes", "check our codebase for CDS adherence", "does this fe...

Why use Cds Code on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/coinbase/cds/tree/master/skills/cds-code. 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 Cds Code?

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 Cds Code?

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

Is the Cds Code AI skill free?

Yes. It is published on GitHub by coinbase under the Apache-2.0 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 👇