跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
设计与 UI

Shadcn UI & Blocks

This skill should be used when the user asks to "build a frontend", "create a landing page", "add shadcn components", "install shadcn blocks", "set up shadcn", "add a hero section", "add pricing section", "build a UI with Tailwind", "use shadcnblocks", "add UI blocks", "add a contact form", "build a dashboard", "create an ecommerce page", "add a navbar", "add a footer", or is building any frontend that could leverage shadcn/ui components and Tailwind CSS. It provides knowledge of 1,338 premium blocks and 1,189 free components from ShadcnBlocks, enabling intelligent selection and installation of the right UI elements for any frontend task.

masonjames · 已有 AI 审核记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add masonjames/Shadcnblocks-Skill --skill Shadcn UI & Blocks

技能说明


name: Shadcn UI & Blocks description: This skill should be used when the user asks to "build a frontend", "create a landing page", "add shadcn components", "install shadcn blocks", "set up shadcn", "add a hero section", "add pricing section", "build a UI with Tailwind", "use shadcnblocks", "add UI blocks", "add a contact form", "build a dashboard", "create an ecommerce page", "add a navbar", "add a footer", or is building any frontend that could leverage shadcn/ui components and Tailwind CSS. It provides knowledge of 1,338 premium blocks and 1,189 free components from ShadcnBlocks, enabling intelligent selection and installation of the right UI elements for any frontend task.

Shadcn UI & ShadcnBlocks Integration

This skill enables building polished frontends using shadcn/ui components and premium ShadcnBlocks templates. Its primary capability is intelligently recommending the best block or component for any UI need — not just the right category, but the right variant within that category — then handling setup and installation.

When to Use

Activate this skill when:

  • Creating a new frontend, landing page, marketing site, or app UI
  • The project uses or should use React/Next.js with Tailwind CSS
  • Pre-built UI sections (hero, pricing, features, testimonials, etc.) would accelerate development
  • The user needs a specific UI element (form, chart, dialog, navigation, etc.)
  • The user mentions shadcn, shadcn/ui, shadcnblocks, or Tailwind-based UI components

Blocks vs Components

BlocksComponents
WhatFull page sectionsReusable UI elements
SizeHero, pricing, footer, etc.Buttons, inputs, dialogs, etc.
Count1,338 across 71 categories1,189 across 60+ groups
AccessPremium (API key required)Free
Install@shadcnblocks/<name><number>@shadcnblocks/<type>-<style>-<number>

Use blocks to compose full page layouts from pre-built sections. Use components for individual UI elements inside custom code.


Intelligent Block Selection

The 0-2 Question Rule

When recommending blocks, follow this progressive disclosure flow:

  • 0 questions: The user's request is specific enough to match directly (e.g., "add a hero with email capture form" → match on cta_pattern: form_capture)
  • 1 question: The section is clear but the style/layout is ambiguous (e.g., "add a feature section" → ask about layout preference)
  • 2 questions max: Both section and style are unclear (e.g., "add a section to show our product" → ask section type, then style)

Never ask more than 2 questions. If still ambiguous, pick the best 2-3 and explain the tradeoffs.

Decision Flow: Infer → Ask → Shortlist

Step 1 — Infer from the request (always do this first)

Extract as many constraints as possible before asking anything:

  • Section type: hero, feature, pricing, testimonial, cta, faq, navbar, footer, contact, stats
  • Goal keywords: "signup" → convert/capture_leads, "trust" → trust, "compare plans" → compare
  • Layout hints: "grid" → grid_cards, "side by side" → split, "accordion" → accordion interaction
  • Tone hints: "clean" → minimal, "bold" → bold, "professional" → enterprise
  • Media hints: "with screenshots" → screenshot media, "video" → video media
  • Density hints: "simple" → minimal density, "detailed" → rich density

Step 2 — Ask targeted questions (only if needed)

Use AskUserQuestion for structured multiple-choice refinement. Consult references/block-index.md → selection_prompts for the pre-defined questions and options per section.

Sections with pre-defined prompts: hero, feature, pricing, testimonial, cta, faq, navbar, footer, contact, stats.

Example for a feature section request:

AskUserQuestion:
  question: "How do you want to showcase your features?"
  options:
    - "Card grid (3-4 columns with icons)" — Classic feature grid with icon + description per card
    - "Alternating image + text rows" — Deep-dive with large visuals, great for product tours
    - "Bento / asymmetric grid" — Modern editorial layout with mixed-size cards
    - "Simple icon list or checklist" — Compact, scannable benefit list

If AskUserQuestion is unavailable, ask the same question inline with numbered options.

Step 3 — Match tags and return 2-3 best picks

Map the user's answer to tag filters, then match against references/block-index.md entries. Return exactly 2-3 recommendations in this format:

Here are the best matches for your [section type]:

1. **[block-id]** — [1-line "why this one"]
   Layout: [layout] · Tone: [tone] · [key differentiator]

2. **[block-id]** — [1-line "why this one"]
   Layout: [layout] · Tone: [tone] · [key differentiator]

3. **[block-id]** — [1-line "why this one"]
   Layout: [layout] · Tone: [tone] · [key differentiator]

Which one works best? I'll install it and set it up.

After the user picks, provide the install command and compose it into the page.

Tag Dimensions Used for Matching

These are the key dimensions from references/block-index.md used to differentiate blocks:

DimensionWhat it tells youExample values
layoutVisual arrangementcentered, split_right_media, grid_cards, bento, alternating
toneAesthetic feelminimal, modern, bold, enterprise, elegant, playful
goalBusiness purposeawareness, explain, trust, convert, capture_leads, compare
cta_patternAction typesingle_primary, primary_secondary, form_capture, multi_action
mediaVisual contentscreenshot, illustration, video, icons, avatars, logos
interactionDynamic behaviorstatic, accordion, tabs, carousel, toggle, hover
items_countNumber of items displayedfew, moderate, many
content_densityHow much contentminimal, stand
阅读完整来源 ↗