Shadcn Ui logo

Shadcn Ui

Community
nahid-sparktales
shadcn-ui

Work with shadcn/ui components and the Radix primitives under them — components are copied into the repo and are your source, so customize them in place, know that re-running the generator overwrites local edits, and know which accessibility behaviour Radix gives you and which it does not. Fires when adding, customizing or debugging a component in a repo with components.json and a vendored UI directory. Not for choosing a component library, not for general component structure, and not an accessibility audit.

Overview

Publishernahid-sparktales
Repositoryagent-dispatcher
Skill nameshadcn-ui
Stars
58
Forks
5
Bundled files
1
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.

  • 1 bundled files

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

  • Open source

    Published by nahid-sparktales on GitHub. Read the source before you install it.

Installation

Install the Shadcn Ui 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/nahid-sparktales/agent-dispatcher/tree/main/skills/frontend/shadcn-ui
  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/nahid-sparktales/agent-dispatcher.git /tmp/agent-dispatcher
mkdir -p .claude/skills
cp -r /tmp/agent-dispatcher/skills/frontend/shadcn-ui .claude/skills/shadcn-ui
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Shadcn Ui 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 Shadcn Ui 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 Shadcn Ui 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.

shadcn/ui and Radix primitives

These components are not a dependency. The generator copied source into the repository, and from that moment it is ordinary project code that nobody upstream will patch for you.

When this fires

Adding, customizing or debugging a shadcn/ui component in a project that has components.json and a vendored UI directory. It does not fire for choosing a component library, for a project using a conventional installed component package, or for a full accessibility audit.

Procedure

  1. Confirm the model before assuming it. components.json, a UI directory of component source, and @radix-ui/* entries in package.json together mean copy-in. Read components.json for the path aliases and conventions the generator will use — it decides where files land.
  2. Check whether the component already exists in the repo before adding it. Re-running the generator for a component that is already there overwrites the file, including every local change. If the file exists and has been customized, adding it again is destructive — stop and ask rather than regenerating over someone's work.
  3. Add via the project's own generator invocation. Take the exact command from the project's README or scripts; the CLI's package name and flags have changed across versions, so do not type one from memory. If the generator cannot be run here, copy the upstream component source into the same directory by hand and add the Radix dependency it imports — do not hand-roll a substitute primitive, and do not silently skip the dependency.
  4. Customize in place. That is the model, not a workaround. Editing the vendored file is the intended way to change these components. Prefer the two low-damage forms: add a variant to the component's existing variant definition, and change appearance through the theme tokens rather than by rewriting markup (see tailwind).
  5. Wrap instead of edit when the change is product-specific. A project-specific composition (your form field, your confirm dialog) belongs in a wrapper around the primitive, leaving the vendored file close to upstream and re-addable. Deep edits to the vendored file are fine, but they are the file you can no longer regenerate cheaply.
  6. Record which vendored files carry local edits. A comment at the top of the file, or a line in the project's own notes — whichever the project already does. Without it, the next person regenerates and loses the change silently.
  7. Keep the component's class-merge path intact. These components merge an incoming className with their own classes through a merge helper; if you rewrite that, caller overrides stop working in ways that look like a styling bug.
  8. Use asChild correctly when composing triggers. It renders the primitive's behaviour onto your child element instead of its own: the child must be a single element that forwards props and ref. A fragment, two children, or a component that drops props produces a control that looks right and does nothing.
  9. Know what Radix gives you and what it does not. Free from the primitive: keyboard interaction and focus management for the pattern, focus trapping and restoration in overlays, dismissal behaviour, correct roles and state attributes, and the internal wiring between a trigger and its content. Not free: an accessible name for an icon-only trigger, a label associated with your own input, error text wired to the field it describes, colour contrast, heading order, honouring reduced-motion, and a visible focus indicator if you removed the ring. Those are yours every time.
  10. Verify with the keyboard and the accessibility tree, not a screenshot — tab to the control, operate it, Escape out of it, and check the name it exposes. Rendered verification is browser-verification.

Checklist

  • components.json read; the component's target path known
  • Existing file checked before any generator run, and overwrite consent obtained if needed
  • Generator command taken from the project, not from memory
  • Customization done in place, preferring variants and theme tokens
  • Product-specific behaviour put in a wrapper, not baked into the primitive
  • Local edits to vendored files recorded
  • className merge path intact; caller overrides still apply
  • Every asChild usage has exactly one prop-and-ref-forwarding child
  • Icon-only controls have accessible names; inputs have associated labels
  • Focus visible, keyboard path walked, Escape/dismiss checked

Failure handling

  • A caller's className does nothing — the merge helper was bypassed or the class conflicts with the component's own. Fix the merge, not with !important.
  • A trigger renders but does not open — suspect asChild with a child that swallows props or ref, or two children where one is required.
  • The component switches between controlled and uncontrolled — a value prop that starts undefined and later becomes defined. Pick one mode and keep it.
  • "Upgrade shadcn" is asked for — there is no upgrade command for code you own. Diff the current upstream source against each vendored file and apply changes deliberately, file by file, keeping local edits. Say which files you did not touch.
  • The generator is unavailable or offline — say so, copy the source manually or stop; do not invent an approximation of a primitive and present it as the component.
  • Accessibility looks handled because Radix is present — it is not a conclusion you can draw from the import. The list in step 9 is checked by hand or reported as unchecked.

Evidence to report

Which components were added or edited and at what paths; whether each vendored file now carries local changes; how customization was done (variant, token, wrapper, direct edit); the keyboard and accessibility-tree checks you actually performed and what they showed; and anything from the step-9 list you did not check.

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 Shadcn Ui AI skill do?

Work with shadcn/ui components and the Radix primitives under them — components are copied into the repo and are your source, so customize them in place, know that re-running the generator overwrites local edits, and know which accessibility behaviour Radix gives you and which it does not. Fires when adding, customizing or debugging a component in a repo with components.json and a vendored UI directory. Not for choosing a component library, not for general component structure, and not an accessibility audit.

Why use Shadcn Ui on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/nahid-sparktales/agent-dispatcher/tree/main/skills/frontend/shadcn-ui. 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 Shadcn Ui?

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 Shadcn Ui?

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

Is the Shadcn Ui AI skill free?

Yes. It is published on GitHub by nahid-sparktales 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 👇