HeroUI MCP Server
Provides AI assistants with access to HeroUI component library documentation, enabling them to list components, retrieve detailed API references, and explore customization options for building React applications.
README
HeroUI MCP Server
A Model Context Protocol (MCP) server providing comprehensive access to the HeroUI component library documentation. Deploy this MCP to help AI assistants like Claude Code build consistent, well-designed React applications using HeroUI components.
🚀 Quick Deploy
This will deploy your HeroUI MCP server to a URL like: heroui-mcp.<your-account>.workers.dev/sse
📋 What This MCP Provides
This MCP server gives AI assistants instant access to HeroUI's 50+ React components with 7 specialized tools:
Available Tools
list_components- Browse all 50 available HeroUI componentsget_component_info- Get component details, imports, and basic usageget_component_api- View complete API reference with props and typesget_component_slots- Explore slot-based customization optionsget_variants_guide- Learn theextendVariantscustomization systemsearch_components- Find components by use case (forms, navigation, overlays, etc.)get_theme_info- HeroUI theme system and configuration guide
Component Categories
- Forms: Input, Textarea, Select, Checkbox, Radio, Switch, Slider, Date Pickers
- Navigation: Navbar, Breadcrumbs, Tabs, Links
- Layout: Card, Divider, Spacer, Accordion
- Data Display: Table, Listbox, Pagination, Avatar, User, Badge
- Feedback: Alert, Toast, Progress, Spinner, Skeleton
- Overlays: Modal, Drawer, Popover, Tooltip, Dropdown
- Media: Image, Avatar
- And more: 50+ components total
🔌 Installation
Option 1: Quick Start (if you don't want to deploy to Cloudflare or run your own locally)
Use the public HeroUI MCP server to get started immediately:
Claude Code:
claude mcp add heroui npx mcp-remote https://heroui-mcp-cloudflare.jgentes.workers.dev/sse
Cursor:
Install in Cursor:
To install the HeroUI MCP Server in Cursor, copy and paste the following link into your browser's address bar:
cursor://anysphere.cursor-deeplink/mcp/install?name=heroui&config=eyJjb21tYW5kIjoibnB4IiwiYXJncyI6WyJtY3AtcmVtb3RlIiwiaHR0cHM6Ly9oZXJvdWktbWNwLWNsb3VkZmxhcmUuamdlbnRlcy53b3JrZXJzLmRldi9zc2UiXX0=
Or manually add to your Cursor settings:
{
"mcpServers": {
"heroui": {
"command": "npx",
"args": [
"mcp-remote",
"https://heroui-mcp-cloudflare.jgentes.workers.dev/sse"
]
}
}
}
Option 2: Deploy Your Own
Deploy your own instance to Cloudflare Workers:
- Click the "Deploy to Cloudflare Workers" button at the top
- After deployment, your MCP will be available at:
https://heroui-mcp-cloudflare.<your-account>.workers.dev/sse - Connect to Claude Code:
claude mcp add heroui npx mcp-remote https://heroui-mcp-cloudflare.<your-account>.workers.dev/sse
Or manually edit your config:
{
"mcpServers": {
"heroui": {
"command": "npx",
"args": [
"mcp-remote",
"https://heroui-mcp-cloudflare.<your-account>.workers.dev/sse"
]
}
}
}
Option 3: Local Server
For running using the Wrangler CLI:
# Install dependencies
npm install
# Run locally (starts on http://localhost:8787)
npm run dev
Connect to your local instance:
Claude Code:
claude mcp add heroui npx mcp-remote http://localhost:8787/sse
Manual config:
{
"mcpServers": {
"heroui": {
"command": "npx",
"args": [
"mcp-remote",
"http://localhost:8787/sse"
]
}
}
}
🛠️ How It Works
This MCP server fetches documentation directly from heroui.com and provides structured information to AI assistants. It helps ensure:
- ✅ Consistency - Correct HeroUI component usage across projects
- ✅ Efficiency - Quick component discovery prevents unnecessary custom development
- ✅ Accuracy - Direct access to official documentation ensures proper implementation
- ✅ Customization - Full support for HeroUI's slots and variants system
- ✅ Best Practices - Includes accessibility guidelines and proper usage patterns
📚 Example Usage
When connected to Claude Code, you can ask questions like:
- "List all available HeroUI components"
- "Show me how to use the Button component"
- "What slots does the Card component have?"
- "Find components for building forms"
- "How do I customize variants in HeroUI?"
The MCP will provide accurate, up-to-date information from the official HeroUI documentation.
🔧 Customization
To modify the MCP server, edit src/index.ts:
- Update component lists in the
COMPONENTSconstant - Add new tools using
this.server.tool(...) - Modify search categories in
COMPONENT_MAP - Add more slot definitions in
SLOTS_INFO
📦 Versioning & Releases
This project uses release-it with conventional changelog for versioning.
For Maintainers
To create a new release:
npm run release
This will:
- Prompt you to select version bump (patch/minor/major)
- Update package.json version
- Generate/update CHANGELOG.md from commit messages
- Create a git tag
- Push changes to GitHub
- Create a GitHub release
Commit Message Format
Use Conventional Commits format for automatic changelog generation:
feat: add new component tool- New features (minor version bump)fix: correct slot parsing- Bug fixes (patch version bump)docs: update installation guide- Documentation changeschore: update dependencies- Maintenance tasksBREAKING CHANGE:in commit body - Breaking changes (major version bump)
See CHANGELOG.md for release history.
📄 License
MIT
🙏 Credits
- Built with Cloudflare Workers
- Documentation from HeroUI
- Implements the Model Context Protocol
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。