
megabytemark/markdown-ui-dsl
Developer toolsこのページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る
A lightweight, text-based UI wireframing schema optimized for AI coding agents and Spec-Driven Development (SDD).
概要
A lightweight, text-based wireframing standard designed specifically for AI coding agents and Spec-Driven Development (SDD). Visual wireframes (Figma, PNGs) are token-heavy, prone to AI hallucination, and hard to version control. Natural language prompts often lack deterministic structure, leading to inconsistent UI generation. This project provides a strict Markdown-based Intermediate Representation (IR). Product managers can write easily readable low-fidelity specs, and AI agents can parse them with deterministic accuracy to generate React, Vue, HTML, or any other UI code. For the smoothest experience, you should integrate the skill file into your AI agent's specific instruction architecture. If your agent supports OpenClaw registry integration, you can install this skill directly via the marketplace. Visit the Markdown-UI DSL page on OpenClaw Hub for one-click installation or use their CLI: Copilot Agent mode works best when using its explicit Skill architecture.
README
以下は原文 README の抜粋です。原文を見る。
Markdown UI DSL for AI Agents
A lightweight, text-based wireframing standard designed specifically for AI coding agents and Spec-Driven Development (SDD).
⭐️ If you find this project useful for your AI tooling, please consider giving it a star! ⭐️
Table of Contents
- The Problem
- The Solution
- Installation & Setup
- Recommended Project Structure
- Workflow
- Syntax Quick Reference
- Design Theming
- Responsive Design
- Events & Interactivity
- Contributing
- License
The Problem
Visual wireframes (Figma, PNGs) are token-heavy, prone to AI hallucination, and hard to version control. Natural language prompts often lack deterministic structure, leading to inconsistent UI generation.
The Solution
This project provides a strict Markdown-based Intermediate Representation (IR). Product managers can write easily readable low-fidelity specs, and AI agents can parse them with deterministic accuracy to generate React, Vue, HTML, or any other UI code.
Installation & Setup
For the smoothest experience, you should integrate the skill file into your AI agent’s specific instruction architecture.
1. OpenClaw Hub (Recommended)
If your agent supports OpenClaw registry integration, you can install this skill directly via the marketplace. Visit the Markdown-UI DSL page on OpenClaw Hub for one-click installation or use their CLI:
claw install markdown-ui-dsl
2. GitHub Copilot (Agent Mode)
Copilot Agent mode works best when using its explicit Skill architecture. Set it up using these steps:
- Initialise your
copilot-instructions.mdby asking Copilot to setup your project for AI agent assistance, or manually create.github/copilot-instructions.md. - Create a folder for the skill and download the DSL instructions:
mkdir -p .github/skills/markdown-ui-dsl
curl -o .github/skills/markdown-ui-dsl/SKILL.md https://raw.githubusercontent.com/MegaByteMark/markdown-ui-dsl/main/skills/markdown-ui-dsl/SKILL.md
- Open your
.github/copilot-instructions.mdfile and add the following mapping:
## Skills
- For interpreting `ui.md` files, use the `markdown-ui-dsl` skill to generate components based on the specifications provided.
2. Cursor / Roo Code / Cline
For agents that use single flat rules files, you can append the skill directly:
# For Cursor (picks up from .cursorrules)
推奨ツール
別のキーワードを試すか、フィルタを外してください。
インストール
npx skillfish add megabytemark/markdown-ui-dsl