huashu-design — Claude Code skill: what it does, how to install
huashu-design builds interactive HTML prototypes, slide decks, and animations right inside your agent. It offers three visual directions first and exports files to PPTX, PDF, MP4, and GIF.
MIT 208 installs ★ 35 updated 2026-10-21
npx skills add podo/design-agent-skills --skill huashu-design -a claude-codeWhat huashu-design does
This skill produces HTML and CSS layouts, clickable prototypes, slide presentations, and data visualizations without relying on UI component frameworks. Before generating files, it drafts three visual directions and pauses until you select one. It then audits the result with design_lint.py and logs the chosen style in a ledger to avoid repeating patterns on later screens.
Core helper utilities require Python 3. Rendering slide decks and videos in PPTX, PDF, MP4, and GIF formats depends on Node.js with playwright, pptxgenjs, sharp, pdf-lib, and ffmpeg configured on your PATH. The repository weighs around 224 MB. The skills CLI may time out during cloning unless you set SKILLS_CLONE_TIMEOUT_MS=900000.
Without ffmpeg on your system PATH, video and GIF exports fail instantly. The skill focuses strictly on HTML layouts, so it does not fit fullstack web applications or mobile frameworks like React Native and SwiftUI. Internal prompts are in Chinese, but upstream notes confirm that English tasks work as expected.
What it helps with
- generate clickable HTML and CSS prototypes directly in chat
- compare three visual design directions before writing code
- export web animations and slide decks to MP4, GIF, PDF, or PPTX
- audit layout markup using the built-in design-lint script
Example prompts for the agent
- Design an interactive landing page prototype using magazine typography with huashu-design
- Create a CSS Grid slide deck and export it to PPTX
- Build an animated prototype for onboarding and render it as an MP4 export
How to install huashu-design
1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill huashu-design -a claude-code2Manually: put the files in place
~/.claude/skills/huashu-design/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/huashu-design/
1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill huashu-design -a codex2Manually: put the files in place
.agents/skills/huashu-design/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill huashu-design -a cursor2Manually: put the files in place
.cursor/skills/huashu-design/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install podo/design-agent-skills skills/huashu-design2Manually: put the files in place
.github/skills/huashu-design/SKILL.md1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill huashu-design -a gemini-cli2Platform CLI
gemini skills install https://github.com/podo/design-agent-skills --path skills/huashu-designManually: put the files in place
.devin/skills/huashu-design/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill huashu-design -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about huashu-design
How to install huashu-design in Claude Code?
Run npx skills add podo/design-agent-skills --skill huashu-design -a claude-code. The skill file installs to ~/.claude/skills/huashu-design/SKILL.md.
What does huashu-design do?
It creates high-fidelity HTML and CSS mockups, animations, and slide decks, offers three visual directions before coding, and exports files to PPTX, PDF, MP4, and GIF.
What dependencies are required for media exports?
Exports require Python 3, Node.js with playwright, pptxgenjs, sharp, and pdf-lib, along with ffmpeg installed on your PATH.
Source text
SKILL.md · MIT
--- name: huashu-design description: | HTML-native design skill for high-fidelity prototypes, slide decks, animations, data visualizations and expert design review, with a library of named design styles and PPTX, PDF, MP4 and GIF export. Agent-agnostic; its own prompts are written in Chinese. triggers: - "huashu" - "magazine typography" - "CSS Grid design" - "animated prototype" - "MP4 export" das: type: skill category: design-systems upstream: "https://github.com/alchaincyf/huashu-design" upstream_path: "SKILL.md" version: latest install: true --- # huashu-design > Catalogue stub — full skill: [alchaincyf/huashu-design](https://github.com/alchaincyf/huashu-design) <!-- das:install:start — generated from stub.yaml by `npm run generate`; edit stub.yaml, not this block --> ## Decision tree 1. **Is the full skill already installed?** - Global: `f=~/.agents/skills/huashu-design/SKILL.md; [ -f "$f" ] && ! grep -q "^das:" "$f" && echo "installed" || echo "pointer"` - Project: `f=.agents/skills/huashu-design/SKILL.md; [ -f "$f" ] && ! grep -q "^das:" "$f" && echo "installed" || echo "pointer"` - `installed` → the skill is installed as `huashu-design`, invoke it and proceed - `pointer` → go to step 2 2. **Detect scope, then install:** ```bash [ -e ~/.agents/skills/huashu-design ] && echo "global" || echo "project" ``` **Global:** ```bash npx skills add alchaincyf/huashu-design --skill huashu-design -g -y ``` **Project:** ```bash npx skills add alchaincyf/huashu-design --skill huashu-design -y ``` > **Claude Code:** send a command as a chat message starting with `!` to run it without leaving the conversation. **Heads-up** - **Requires:** Python 3 for the skill's helper scripts (learn.py, design_lint.py, capabilities.py); Node.js with the packages in the skill's package.json (Playwright, pptxgenjs, sharp, pdf-lib) and ffmpeg on PATH for PPTX, PDF, MP4 and GIF export; the skills CLI installs none of these - **License:** MIT - **Large download (~224 MB):** the skills CLI gives up cloning after 300 s; if it times out, retry with `SKILLS_CLONE_TIMEOUT_MS=900000`. <!-- das:install:end --> ## Invoke after install - Skill name: `huashu-design` - Trigger phrases: "huashu", "magazine typography", "CSS Grid design", "animated prototype", "MP4 export" ## What it does Produces publication-quality HTML/CSS designs: clickable prototypes, slide decks, animations, infographics and visualizations, with expert design review. For any new design it first shows three directions and waits for your pick before executing, draws on a library of named design styles with visual specimens, runs a design-lint check on the delivered file, and keeps a "look ledger" of what you have shipped so later projects avoid repeating it. Supports animated interactive prototypes and exports to PPTX, PDF, MP4 and GIF. Agent-agnostic and does not need an external component library. The skill's prompts and references are in Chinese, but upstream says it works fine with English tasks. ## When NOT to use - Production web apps or systems that need a backend — upstream says the skill does not apply - Non-HTML targets such as React Native, SwiftUI, or backend rendering (the skill is HTML/CSS native) - When you need a pre-built component library or design system scaffold (use `shadcn-ui` or `design-consultation`)
© podo · MIT · podo/design-agent-skills