mcp-interactive-ui-server
Enables AI agents to render interactive user interfaces such as forms, dashboards, charts, tables, and wizards directly in MCP-compatible clients. Supports structured data collection and richer interactions beyond text responses.
README
mcp-interactive-ui-server
<!-- mcp-name: mcp-interactive-ui-server -->
MCP Apps UI toolkit — forms, dashboards, charts, tables for AI agents to render interactive UIs.
What Are MCP Apps?
MCP Apps is a 2026 extension to the Model Context Protocol that enables AI agents to go beyond text responses and render interactive user interfaces directly within MCP-compatible clients.
Instead of returning plain text or markdown, agents can generate structured UI schemas that clients render as native components — forms, dashboards, charts, tables, and multi-step wizards. This bridges the gap between conversational AI and application-level interactivity.
Why This Matters
- Agents become app builders: Instead of describing data, agents can present it visually with charts, dashboards, and tables
- Structured data collection: Forms and wizards let agents gather complex user input with validation
- Richer interactions: Toggles, buttons, accordions, and tabs embedded directly in responses
- Client-agnostic: The JSON schema approach works across any MCP client that supports UI rendering
Tools (6)
| Tool | Description |
|---|---|
create_form |
Generate interactive forms with text, number, select, checkbox, date fields and validation |
create_dashboard |
Build grid-based dashboards with stat cards, charts, tables, progress bars |
create_chart |
Define charts (bar, line, pie, scatter, area, donut, heatmap) with data and styling |
create_table |
Create sortable, filterable, paginated tables from headers and row data |
create_wizard |
Multi-step form wizards with per-step validation and progress tracking |
render_markdown |
Enhanced markdown with embedded interactive elements (buttons, toggles, inputs, tabs) |
Installation
pip install mcp-interactive-ui-server
Quick Start
Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"interactive-ui": {
"command": "interactive-ui-server"
}
}
}
Cursor / Windsurf
Add to your MCP settings:
{
"mcpServers": {
"interactive-ui": {
"command": "interactive-ui-server"
}
}
}
With uvx (no install needed)
{
"mcpServers": {
"interactive-ui": {
"command": "uvx",
"args": ["mcp-interactive-ui-server"]
}
}
}
Usage Examples
Create a Contact Form
Agent: create_form("Contact Us", [
{"label": "Name", "type": "text", "required": true},
{"label": "Email", "type": "email", "required": true},
{"label": "Subject", "type": "select", "options": ["General", "Support", "Sales"]},
{"label": "Message", "type": "textarea", "rows": 6}
])
Build a Sales Dashboard
Agent: create_dashboard("Q1 Sales", [
{"type": "stat_card", "title": "Revenue", "value": "$142K", "trend": "up", "change": "+12%"},
{"type": "stat_card", "title": "Orders", "value": "1,247", "trend": "up", "change": "+8%"},
{"type": "chart", "title": "Monthly Sales", "chart_type": "bar",
"data": {"labels": ["Jan","Feb","Mar"], "datasets": [{"label": "Sales", "values": [42,51,49]}]}},
{"type": "progress_bar", "title": "Quarterly Target", "value": 72, "max": 100}
])
Generate a Chart
Agent: create_chart("pie",
{"labels": ["Chrome", "Firefox", "Safari", "Edge"], "values": [65, 15, 12, 8]},
title="Browser Market Share"
)
Interactive Markdown Report
Agent: render_markdown(
"# Status Report\n\nSystem health: {{status_badge}}\n\n{{details}}\n\n{{action_btn}}",
[
{"type": "badge", "id": "status_badge", "text": "Healthy", "color": "green"},
{"type": "accordion", "id": "details", "title": "View Details", "content": "All 12 services operational."},
{"type": "button", "id": "action_btn", "label": "Run Diagnostics", "action": "diagnose"}
]
)
UI Schema
All tools return structured JSON following the mcp-ui schema specification:
{
"$schema": "mcp-ui/<component>/v1",
"type": "<component_type>",
"title": "...",
"...": "component-specific properties",
"metadata": {
"created_at": "ISO timestamp",
"version": "1.0"
}
}
Schema types: form, dashboard, chart, table, wizard, enhanced_markdown
Requirements
- Python 3.10+
- MCP SDK (
mcp[cli]>=1.0.0)
License
MIT
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。