better-design
Open-source design MCP server and shadcn/ui registry: every Better Design design system, with design tokens, UI principles and accessibility rules for Claude Code, Cursor, Codex and any MCP client. Browse the catalog at better-design.com/design-systems.
Runssource: GitHubhomepageTypeScriptMITcommit e146625f3041
TypeScript, MIT licensed. The project labels itself: ai design, claude code, claude code for design, claude design, codex design, coding agents, component library and cursor.
better-design runs. An Argusic agent installed it in 27 minutes and hit 2 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.
The registry repo builds and passes its suite (22 tests, typecheck clean, generator idempotent), and the published better-design CLI is installed with a live free API key that the hosted MCP server accepted, with the MCP server wired into ~/.codex/config.toml.
What the agent ended up with on a clean machine, in its own words. How this is measured
Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 100 of 100 (the mean of this project's run scores).
At a glance
Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.
What was measured
From the run that decided this verdict. Every line is a recorded value, not an estimate.

What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- tests/vitest 5 fails to start on the container's Node 18.19.1 (SyntaxError: node:util does not provide an export named 'styleText')9 minutes
- bun is not installed in the container (documented test runner)8 minutes
- Install time
- 27 minutes
- Cold machine to finish
- 3 minutes
- Errors hit and fixed
- 2 hit, 2 fixed with no human help
- How the result was proved
- cd tests && bun run test -> 'Test Files 2 passed (2), Tests 22 passed (22)'; cd /work/repo && node scripts/generate-transitions.mjs -> 'Wrote 9 transition components + 1 meta entry to 31 design systems' with git status clean; npx -y [email protected] --help exited 0; npx -y [email protected] ask 'List the design systems Better Design supports' exited 0 and returned the live UI-principles document; npx -y [email protected] doctor -> 'OK API key: found in ~/.better-design/key' and 'OK Server: https://better-design.com/api/mcp accepted the key'
- Model tokens used
- 143,113
- Exact commit tested
- e146625f3041
- Container image
- sha256:cdd920bce783
What was not measured
The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.
- Peak GPU memory
- this ran on a CPU machine, so there was no GPU to sample
- Behaviour over a long run
- each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
-rw-rw-r-- 1 runner runner 2235 Oct 19 22:18 /home/runner/.codex/AGENTS.md <!-- better-design:start --> ## Better Design For every UI build, run these steps in order, and never present a page that skipped step 2 or 4: 1. Design system: keep the product's own design. Use its theme and components when the code has the… backups key 41 /home/runner/.better-design/key Everything has been verified. Final state is solid: install, tests (22/22), generator idempotency, … tokens used 143,113 Everything has been verified. Final state is solid: install, tests (22/22), generator idempotency, …
Replay the whole session, every command from a clean machine to this point.
How it was tested
One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.
Strengths and limits
Measured facts, not opinions. How this is written.
What went well
- Reached a running state on a clean machine, with the session recorded.
- Installed in 27 minutes.
- Recovered from all 2 errors without a human stepping in, which says the failures are documented well enough to solve.
- Ran without a GPU, so it does not need one to start.
- Licensed MIT, as reported by its host.
What did not, or is not known
- Measured once so far. A single attempt shows it can run, not that it runs reliably.
What it is a reasonable choice for
- Trying it on a laptop or a small server: it reached a running state without a GPU.
Also tested, in the same area
Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.
Run history
Topics (from GitHub)
ai-designclaude-codeclaude-code-for-designclaude-designcodex-designcoding-agentscomponent-librarycursorcursor-designdesign-systemsdesign-tokensfigma-alternativegenerative-uigithub-copilotmcp-servershadcnshadcn-uitailwindcssui-componentsvibe-coding
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/better-design)Questions
- Does better-design run?
- Yes. better-design runs. Argusic installed and launched it on a clean machine in 27 minutes, hitting 2 errors on the way, and recorded the session.
- How did Argusic test better-design?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit e146625f3041. One attempt is recorded, and the full method is on the methodology page.
- What did it cost to test better-design?
- The run that produced this verdict cost $0.03: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does better-design take to install?
- 27 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
- Does better-design need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing better-design?
- 2 things broke in the recorded run, and 2 were fixed without human help. Each one, and the time it cost, is listed on this page.
- Where is the evidence for better-design?
- The recorded run is on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.