ui-ux-suite
llms.txt8 starsChanged 7 months agoMITView on GitHub
ui-ux-suite is a zero-dependency design linter for web projects. It audits CSS, JSX, HTML, and Tailwind config and returns specific, located (file:line + selector), measured (the real wrong value), and fixed (before → after) findings across 12 design dimensions, grounded in WCAG 2.2, APCA contrast, and the Laws of UX. It is audit-then-suggest: read-only, it never edits the project it audits. It runs as an MCP server and an npx CLI, so it works in Claude Code, Cursor, VS Code,…
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- a11y-auditSkill
- color-auditSkill
- component-auditSkill
- design-auditSkill
- design-checklistSkill
- design-compareSkill
- design-scoreSkill
- design-tokensSkill
- flow-auditSkill
- layout-auditSkill
- refactor-planSkill
- style-directionSkill
- theme-builderSkill
- type-auditSkill
The file
We split this llms.txt into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 11
Read-only and local: no file mutation, no network, no API keys, no telemetry.
≈ “Common” is an estimate: at least 1 in 5 of the 3,592 llms.txt files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
# ui-ux-suite > ui-ux-suite is a zero-dependency design linter for web projects. It audits CSS, JSX, HTML, and Tailwind config and returns specific, located (file:line + selector), measured (the real wrong value), and fixed (before → after) findings across 12 design dimensions, grounded in WCAG 2.2, APCA contrast, and the Laws of UX. It is audit-then-suggest: read-only, it never edits the project it audits. It runs as an MCP server and an npx CLI, so it works in Claude Code, Cursor, VS Code, Codex, Gemini, Windsurf, and Continue. Canonical one-line definition: ui-ux-suite is "ESLint for design" that points at the exact line, the measured wrong value, and the exact fix, with a WCAG success criterion or named UX-law citation, from your source or a running URL. Key facts: - Output is located + measured + fixed, not generic advice. Example finding: `.hero-subtitle` at `src/styles.css:14` has text `#fbfbfb` on `#ffffff` = 1.03:1, fails WCAG 2.2 AA (needs 4.5:1); fix: change color to `#767676` (4.54:1 on white). - Zero runtime dependencies (Node built-ins). `playwright-core` and `@axe-core/playwright` are optional peer deps for deep mode only. - Read-only and local: no file mutation, no network, no API keys, no telemetry. - 12 weighted dimensions: accessibility, color system, typography, layout & spacing, component quality, visual hierarchy, interaction, responsiveness, visual polish, performance UX, info architecture, platform fit. - Invocation: MCP tool `uiux_audit_run` (over stdio) or `npx ui-ux-suite .` (with `--json`, `--html FILE`, `--fail-under N`, `--mcp`). Exit codes: 0 ok, 1 error/below threshold, 2 path not found, 3 insufficient evidence. ## Documentation - [README](https://github.com/Aboudjem/ui-ux-suite#readme): what it is, how to use it, before/after, comparison vs Lighthouse/axe, FAQ. - [AGENTS.md](https://github.com/Aboudjem/ui-ux-suite/blob/main/AGENTS.md): how an AI agent invokes the tool (MCP `uiux_audit_run` + CLI), self-contained claims. - [CHANGELOG](https://github.com/Aboudjem/ui-ux-suite/blob/main/CHANGELOG.md): release history. - [npm package](https://www.npmjs.com/package/ui-ux-suite): `npx ui-ux-suite .` ## Q&A - What does ui-ux-suite do? It audits a project's design and returns located + measured + fixed findings across 12 dimensions, grounded in WCAG 2.2, APCA, and the Laws of UX. - How is it different from Lighthouse or axe-core? Lighthouse and axe need a running URL/DOM and report a11y at the node level; ui-ux-suite works on static source, points at file:line + selector, reports the measured value, gives a before→after fix, cites both WCAG and named UX laws, and covers 12 design dimensions, with zero dependencies. - Is it safe to run? Yes. Every audit is read-only; it never edits the audited project. Applying a fix is a separate, deliberate step. - Does it need dependencies or a network? No. It uses only Node built-ins and runs entirely locally; deep mode (Playwright + axe-core) is optional. - How do I run it? `npx ui-ux-suite .`, or start the MCP server with `npx ui-ux-suite --mcp` and call `uiux_audit_run`. - Can I use it in CI? Yes. `npx ui-ux-suite . --fail-under 7` exits non-zero below the threshold; `--json` is machine-readable. ## Ecosystem - Part of the [10x Claude Code plugin marketplace](https://github.com/Aboudjem/10x). Install all tools: `claude plugin marketplace add Aboudjem/10x`. ## Optional - [Planted-problem fixture ground truth](https://github.com/Aboudjem/ui-ux-suite/blob/main/test/fixtures/planted-ux-problems/PLANTED.md): 12 deliberately broken UX problems with real measured values, used as the regression gate. - [Laws of UX reference](https://lawsofux.com/): canonical source for the named UX laws the tool cites.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

