agentsclimarketplace

Ui ux pro max

Skill sidiangongyuan/codex-skills-library/skills/ui-ux-pro-max

Use when designing, implementing, reviewing, or improving web and mobile UI/UX. Consults searchable design data for product patterns, styles, palettes, typography, accessibility, charts, icons, and common frontend stacks.From its SKILL.md

Install
npx -y skills add sidiangongyuan/codex-skills-library --skill ui-ux-pro-max

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

7.9 KB, ~1.7k tokens by cl100k_base, as published. Nobody here has run it

UI/UX Pro Max

Use the bundled search data to make product-specific interface decisions, then validate the implemented experience. Treat search results as design evidence, not as a substitute for reading the existing product, audience, and codebase.

Use this skill when

  • designing or refactoring a page, screen, component, navigation flow, form, table, chart, design system, or responsive layout;
  • choosing or auditing typography, color, spacing, hierarchy, motion, icons, interaction states, accessibility, or platform conventions;
  • implementing a frontend in a supported stack; or
  • diagnosing an interface that feels unclear, inconsistent, inaccessible, or visually unfinished.

Skip it for backend-only, infrastructure-only, database-only, or non-visual automation work. If a change affects how a feature looks, feels, moves, or is operated, this skill is relevant.

Workflow

1. Understand the product before choosing a style

Inspect the existing interface and code when present. Establish:

  • product type and primary user goal;
  • audience, usage context, and accessibility needs;
  • current design system, component library, icon set, and frontend stack;
  • information density, target devices, themes, and required states; and
  • brand constraints and the intended visual tone.

Prefer the repository's established patterns over a new visual language unless the task explicitly calls for a redesign. Operational products should optimize scanning, comparison, and repeated action; branded or editorial experiences can support a more expressive composition.

2. Generate a design-system recommendation

Start with a multidimensional query: product + industry + tone + density. Run the bundled script from any working directory by resolving <skill-dir> to this installed skill directory:

python "<skill-dir>/scripts/search.py" "healthcare operations dashboard calm dense" --design-system -p "Care Console"

Use python3 instead of python where that is the available executable. The result combines product, style, color, landing, and typography data with reasoning rules and anti-patterns.

Do not apply the first recommendation mechanically. Re-query when it conflicts with the product, established UI, platform conventions, or accessibility.

3. Search only the domains needed for the task

python "<skill-dir>/scripts/search.py" "keyboard focus error recovery" --domain ux -n 5
python "<skill-dir>/scripts/search.py" "financial trend comparison" --domain chart -n 5
python "<skill-dir>/scripts/search.py" "table virtualization rerender" --stack react -n 5
python "<skill-dir>/scripts/search.py" "data table toolbar" --stack shadcn -n 5

Available domains:

DomainUse for
productProduct-specific patterns and reasoning
styleVisual languages, effects, and anti-patterns
colorProduct and industry palettes
typographyFont pairings and type-system direction
google-fontsIndividual Google Font discovery
iconsIcon families, semantics, and consistency
landingLanding-page structure and conversion patterns
chartData relationships, chart types, and libraries
uxAccessibility, interaction, responsive, and state guidance
webWeb and app-interface implementation guidance
reactReact and Next.js performance patterns

Available stacks are react, nextjs, vue, svelte, astro, swiftui, react-native, flutter, nuxtjs, nuxt-ui, html-tailwind, shadcn, jetpack-compose, threejs, angular, laravel, and javafx.

Use --json when another script will consume results. Use -f markdown for human-readable design-system output.

4. Persist decisions only when useful

For a multi-page product or a workflow that will continue across sessions:

python "<skill-dir>/scripts/search.py" "B2B analytics precise compact" --design-system --persist -p "Metrics Hub" --output-dir .

This creates design-system/MASTER.md. A page-specific call adds an override:

python "<skill-dir>/scripts/search.py" "B2B analytics settings form" --design-system --persist -p "Metrics Hub" --page settings --output-dir .

When implementing a page, read the master first, then design-system/pages/<page>.md if it exists. Page rules override the master only where they differ. Do not persist generated files unless they belong in the user's project.

5. Turn evidence into a coherent interface

Define a compact decision set before implementation:

  • content hierarchy and the primary action;
  • layout grid, container behavior, responsive breakpoints, and fixed regions;
  • semantic color, typography, spacing, radius, elevation, and motion tokens;
  • component states: default, hover/press, focus, selected, disabled, loading, empty, error, success, and permission-limited;
  • navigation, keyboard, touch, and screen-reader behavior; and
  • what must be verified in a browser, simulator, or device.

Use familiar symbols or the existing icon library for standard actions. Use real product imagery when the user needs to inspect a product, place, object, person, or state. Avoid decorative visuals that compete with operational work.

6. Validate the implementation

Inspect the actual rendered experience, not only source code. Use the relevant items in references/quality-checklist.md, and query the bundled data when a rule needs more detail.

At minimum verify:

  • representative desktop and mobile widths, including the narrowest supported viewport and landscape where relevant;
  • keyboard order, visible focus, labels, roles, contrast, reduced motion, and enlarged text;
  • loading, empty, error, retry, disabled, overflow, and long-content states;
  • touch targets, safe areas, fixed-element offsets, and absence of unintended horizontal scrolling;
  • theme parity when both light and dark modes exist; and
  • layout stability, media dimensions, and performance on large lists or data views.

For 3D or canvas work, verify that pixels render, the scene is correctly framed, assets load, motion is active where expected, and controls do not overlap the scene across supported viewports.

Priority order

When findings compete, resolve them in this order:

PriorityCategoryNon-negotiable signal
1AccessibilityOperable and understandable without excluding users
2Interaction safetyReliable targets, feedback, recovery, and confirmation
3Information architectureClear hierarchy, navigation, and primary task
4Responsive layoutNo clipping, overlap, hidden content, or accidental scroll
5PerformanceStable layout and responsive input
6Product fitStyle and density match the domain and audience
7Typography and colorReadable, semantic, consistent tokens
8Forms and system statesComplete feedback and error recovery
9MotionMeaningful, interruptible, reduced-motion compatible
10PolishIcons, charts, spacing, and details reinforce the system

Visual novelty never outranks comprehension, accessibility, or task success.

Output expectations

For a design task, return the chosen direction, key tokens and interaction decisions, component or page structure, and validation plan. For an implementation task, make the scoped code changes and report the tested states and viewports. For a review, lead with concrete findings tied to the rendered UI or code and separate blocking usability problems from optional polish.

Do not present database recommendations as universal rules, invent brand requirements, or claim a UI is verified without inspecting the relevant rendered states.

What ships with it: 39 files

1722.3 KB alongside SKILL.md, 4 of them executable

agents/

references/

scripts/

Gives 0 of the 12 instructions most design frontend skills give in ~1.7k tokens

Counted across 1,169 of the 1,878 authors here whose files we hold, read 2026-08-07

  • Use CSS variables for color consistencyin 72 of 1169, across 23 files
  • Commit to one bold aesthetic direction before codingin 72 of 1169, across 27 files
  • Match implementation complexity to the aesthetic visionin 70 of 1169, across 20 files
  • Add atmospheric background effects and texturesin 57 of 1169, across 9 files
  • Use unexpected spatial compositions and layoutsin 56 of 1169, across 8 files
  • Implement real working codein 55 of 1169, across 7 files
  • Vary themes and aesthetics across different designsin 48 of 1169, across 7 files
  • Launch chromium in headless modein 47 of 1169, across 4 files
  • Close the browser when donein 47 of 1169, across 4 files
  • Run provided scripts with help flag firstin 47 of 1169, across 4 files
  • Wait for network idle statein 47 of 1169, across 4 files
  • Use descriptive selectors for elementsin 47 of 1169, across 4 files

Said here and by no other author read

  • Inspect existing interface and code before choosing a style
  • Prefer established repository patterns over new visual languages
  • Generate a multidimensional design-system recommendation query
  • Run the bundled search script to gather design data
  • Re-query when initial recommendations conflict with the product
  • Search only the specific domains needed for the task

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,750. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.