Generative UI for LLMs using Markdoc {% %} tag syntax inline with markdown prose.
LLMs write natural markdown and drop interactive components — charts, buttons, forms, tables — all in the same stream. No custom DSL, no JSON blocks, no JSX confusion.
Shimmer placeholders, Zod prop validation, contextData passthrough, error boundaries, and verbosity prompt modes — all built in.
The Q4 results show strong growth.
{% chart type="bar" labels=["Jan","Feb","Mar"] values=[120,150,180] /%}
Revenue grew **12%**. Want to dig deeper?
{% button action="continue" label="Show by region" /%}
Packages
| Package | Description |
|---|---|
@mdocui/core |
Streaming parser, component registry, system prompt generator |
@mdocui/react |
React renderer, 24 theme-neutral components, useRenderer hook |
@mdocui/cli |
Scaffold, generate system prompts, preview |
Get started
pnpm add @mdocui/core @mdocui/react
For AI coding agents
Install the mdocUI skill in Claude Code:
mkdir -p .claude/skills/mdocui curl -o .claude/skills/mdocui/SKILL.md https://raw.githubusercontent.com/mdocui/mdocui/main/SKILL.md
Then invoke with /mdocui in any conversation.
Links
- Documentation | Live Demo | Examples | SKILL.md | llms.txt