惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
人人都是产品经理
人人都是产品经理
IT之家
IT之家
T
The Blog of Author Tim Ferriss
V
V2EX
博客园 - 聂微东
The Cloudflare Blog
Blog — PlanetScale
Blog — PlanetScale
A
About on SuperTechFans
U
Unit 42
Vercel News
Vercel News
L
LangChain Blog
博客园 - 司徒正美
H
Help Net Security
Recent Announcements
Recent Announcements
Recorded Future
Recorded Future
V
Visual Studio Blog
Jina AI
Jina AI
Microsoft Azure Blog
Microsoft Azure Blog
GbyAI
GbyAI
Y
Y Combinator Blog
C
Check Point Blog
博客园 - 三生石上(FineUI控件)
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
J
Java Code Geeks
The Register - Security
The Register - Security
The GitHub Blog
The GitHub Blog
B
Blog RSS Feed
F
Fortinet All Blogs
B
Blog
G
Google Developers Blog
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
云风的 BLOG
云风的 BLOG
爱范儿
爱范儿
MongoDB | Blog
MongoDB | Blog
F
Full Disclosure
有赞技术团队
有赞技术团队
罗磊的独立博客
博客园_首页
MyScale Blog
MyScale Blog
aimingoo的专栏
aimingoo的专栏
Google DeepMind News
Google DeepMind News
M
MIT News - Artificial intelligence
N
Netflix TechBlog - Medium
Engineering at Meta
Engineering at Meta
量子位
I
InfoQ
小众软件
小众软件
P
Proofpoint News Feed

Open Design Blog

Open Design 0.15.1 — Sharper Vision, Longer Flow Open Design 0.15.0: cost less, ship faster Claude PPT Skills: How to Make a Slide Deck with Your Coding Agent codex-ppt-skill: Image-Forward PPT Decks with a Coding Agent frontend-slides: Make Web Slides with a Coding Agent dashiAI-ppt-skill: Editable Presentations From a Coding Agent guizang-ppt-skill: Editorial HTML Decks with a Coding Agent Marp: The Markdown Presentation Tool (and How to Export to PowerPoint) ppt-master: Editable PowerPoint From Any Document, Built by an Agent reveal.js: Themes, Export & How It Compares to the Alternatives Slidev: Themes, Examples & How to Export to PowerPoint Open Design 0.14.0: the inspiration time machine Open Design AI Workshop is coming to Shanghai Open Design 0.13.0: stay in flow Best AI Design Agents in 2026: An Honest, Tested Guide Best AI Prototyping Tools in 2026: An Honest, Tested Guide Best AI Design Tools in 2026: An Honest, Tested Guide Best Bolt.new Alternatives in 2026: An Honest Comparison Best Design-to-Code Tools in 2026: An Honest, Tested Guide Best Figma Alternatives in 2026: An Honest, Tested Guide How to Use Claude Code for Frontend Design (2026 Guide) Best Lovable Alternatives in 2026: An Honest Comparison Best v0 Alternatives in 2026: An Honest Comparison Open Design is coming to Osaka / Kyoto Open Design 0.12.0: your brand is a design system Vibe Design Tools: An Honest Guide to What Works Vibe Design vs Vibe Coding: Where They Split and Why It Matters Vibe Design with Google Stitch: What It Nails, Where It Traps You What Is Vibe Design? The 2026 Guide to Designing by Intent Open Design 0.11.0: the Bazaar Open Design 0.10.0: the all-in-one design workspace Open Design 0.9.0: design for everyone The open-source alternative to Figma Open Design 0.8.0: everything is a plugin The layout layer the canvas used to hide How to port a Figma workflow into an Open Design plugin BYOK reality check: 5 things that break in Open Design today The open-source alternative to Claude Design 31 skills, 72 systems: how the Open Design library works BYOK design workflow: run Claude, Codex, or Qwen on your own key
Why we built Open Design as a skill layer, not a product
Open Design · 2026-05-13 · via Open Design Blog

Most AI design tools try to replace the agent already on your laptop. Open Design takes the opposite bet: ship a thin layer of skills, systems, and adapters that turn any coding agent into a design engine — without locking you into a new app.

The strongest coding agent on your laptop right now is Claude, Codex, Cursor, Gemini, OpenCode, or Qwen. We don’t think you need another one. What’s missing is not raw intelligence — it’s taste, structure, and a workflow that respects design as a craft.

Open Design is our attempt at that missing layer. It’s not a chat product. It’s not a design tool that “uses AI under the hood.” It’s a thin skill layer — a folder of SKILL.md files, a portable library of design systems, and a daemon that auto-detects your existing CLI agents and wires them together.

This post explains why we made that choice, and what it implies for how you’ll use Open Design.

A product would be the wrong shape

The instinct, when starting an AI design project in 2026, is to build a new app: a chat interface, a canvas, a billing system, a model bill that grows linearly with your user count.

We considered that path and rejected it for three reasons:

  1. The chat interface is a commodity. Every user already has one. Adding a worse one helps nobody.
  2. The model bill is a tax on creativity. Bundling inference into the product means we have to mark it up, throttle long sessions, and ration model upgrades. None of that serves the designer.
  3. Lock-in is the wrong default. Designers should be able to leave with their files, their systems, their skills. A product wraps everything in proprietary state. A skill layer doesn’t.

So we built the layer instead. Drop a folder, restart the daemon, the skill appears. Take the folder with you, drop it into a different agent, the skill works there too.

What a skill actually is

A skill in Open Design is a SKILL.md file plus optional supporting assets in the same folder. The Markdown file describes:

  • What the skill does — one paragraph, in plain English
  • When to invoke it — the trigger conditions, written so the agent can route correctly
  • The shape of the output — HTML, PDF, slides, a Markdown brief
  • The constraints — palette in OKLch, font stack, layout posture, brand vocabulary

The agent reads the file, decides whether to invoke, and writes the output to disk. There is no plugin system, no API surface, no version compatibility matrix. If you can write Markdown, you can ship a skill.

This is intentional. We’ve watched plugin ecosystems decay for fifteen years — each one a trade between expressiveness and longevity, won by neither. Files are forever. Plugins are a snapshot of someone’s API in 2024.

Why systems are also Markdown

Open Design ships 72 design systems — Linear, Vercel, Stripe, Apple, Cursor, Figma, and more — as DESIGN.md files. Same idea: portable, readable, agent-ingestible.

A design system, in this context, is not a Figma library. It’s a contract: here are the colors in OKLch, here is the type ramp, here is the spacing scale, here is the layout posture, here is the voice. The agent reads the contract and produces work that respects it.

You can fork a system, edit it in place, ship a variant. You can write your own from scratch in 30 minutes. You can mix systems mid-project — pull the typography from Linear, the color logic from Vercel, the layout from a custom in-house spec — because they’re all just text.

BYOK is the only honest model

Open Design runs on bring-your-own-key. You paste a base URL and an API key for any OpenAI-compatible endpoint — DeepSeek, Groq, OpenRouter, your own self-hosted vLLM — and you’re done.

We don’t run inference. We don’t take a margin on tokens. We don’t have a billing relationship with you. That’s not a sustainability problem — it’s the only honest answer to the question “who pays when the agent runs?”

The answer is: you do, directly, to the model provider you chose. We get out of the way.

What this means for you

If you want a polished SaaS with a nice chat box and a single subscription, we’re not the right tool. There are good products in that shape — use them.

If you want a workflow where:

  • the agent you already trust does the work,
  • the skills are files you can read and edit,
  • the design systems are portable across projects and agents,
  • and the bill goes to the model provider, not us —

then Open Design is built for you. Drop into the GitHub repo, run pnpm tools-dev, point your agent at a skill, and ship.

The skill layer wins because it doesn’t compete with the agent on your laptop. It augments it.