nextlevelbuilder/ui-ux-pro-max-skill
An AI skill that provides design intelligence for building professional UI/UX across multiple platforms.
How ui-ux-pro-max-skill is put together
This repository packages an AI 'skill' that provides UI/UX design intelligence (192 reasoning rules, 79 UI styles, color palettes, typography pairings) as CSV/JSON knowledge bases consumed by AI coding assistants. It is a multi-workspace monorepo: a TypeScript/Node CLI (`cli/`) published to npm as `ui-ux-pro-max-cli` that installs the skill assets into projects, a Next.js 14 'Style Gallery' web app (`gallery/`) for browsing styles, a Claude Code 'stack' bundle (`stack/`), and Python validation/evaluation scripts. Data flows from CSV knowledge files (`src/ui-ux-pro-max/data/`, mirrored into `cli/assets/data/`) through a BM25-style reasoning/search process into generated design-system recommendations. The primary implementation languages are TypeScript (CLI and gallery) and Python (data validation/evaluation tooling).
Languages
Frameworks
Datastores
Infrastructure
Major components
cli
TypeScript Node CLI (`uipro`) that installs, updates, and uninstalls the UI/UX Pro Max skill into AI coding assistant projects.
gallery
Next.js 14 + React web app that renders a searchable/filterable gallery of UI styles with previews, color palettes, and code snippets.
ui-ux-pro-max data (src/ui-ux-pro-max/data)
Canonical CSV/JSON knowledge base (styles, colors, typography, products, ui-reasoning, ux-guidelines) powering the design intelligence.
cli/assets/data
Bundled mirror of the knowledge-base CSV/JSON assets shipped inside the CLI package for offline template installs.
.claude/skills
Claude Code skill definitions (SKILL.md plus reference/template data) for design, brand, design-system, slides, and ui-styling.
stack
Ready-to-use Claude Code stack bundling the skill with agents, commands, MCP config, and a Playwright-based design-audit script.
python scripts (validation/evaluation)
Python3 tooling that validates CSV/semantic data, runs relevance evaluation, and executes unit tests over the knowledge base.
projects/preview samples
Static HTML example outputs (healthcare-dashboard, portfolio-dark, saas-landing, preview app) demonstrating generated designs.
Broadened AI assistant support and shored up test coverage for script paths.
5 weeksWeek of 2026-08-24
Week of 2026-08-24latest
Broadened AI assistant support and shored up test coverage for script paths.
Week of 2026-08-17
Focused on documentation quality, including a new Vietnamese translation.
Week of 2026-08-10
Overhauled search relevance and clarified several documentation points.
Week of 2026-08-03
Automated version bumping and added flexible installation options.
Week of 2026-07-27
A busy week expanding assistant compatibility and fixing installation and design-system behavior.
Over the past month the project matured from adding core features to polishing and broadening its reach. Early weeks focused on expanding support for many AI coding assistants and fixing how the tool installs and runs scripts, while later weeks shifted toward documentation, translations, and search-quality improvements. The overall arc shows a maturing tool stabilizing its command-line experience and welcoming a wider, more international audience.
Week by week
2026-08-24Broadened AI assistant support and shored up test coverage for script paths.latest2 changes
Docs
Added OpenClaw to the assistant list
The README now lists OpenClaw as one of the AI coding assistants the tool can set up.
Chore
Added Droid to script-path checks
Automated tests now confirm the tool generates correct file paths for the Droid assistant.
2026-08-17Focused on documentation quality, including a new Vietnamese translation.3 changes
Docs
Added a Vietnamese README
Vietnamese-speaking users can now read the project's main documentation in their own language.
Docs
Corrected outdated contributor instructions
Fixed stale manual-sync steps in the contributing guide so new contributors follow the right process.
Fix
Restored the star-history chart
Repaired a broken chart in the docs that shows the project's growth in GitHub stars over time.
2026-08-10Overhauled search relevance and clarified several documentation points.4 changes
Feature
Improved search relevance and design data
Reworked how the tool ranks results and curated its design data to return better matches.
Fix
Fixed generated script paths
The command-line tool now produces script paths that actually run as expected.
Docs
Clarified the asset sync workflow
Documentation now more clearly explains how project assets are kept in sync.
Docs
Clarified Claude ZIP upload limit
Added a note explaining the size limit when uploading ZIP files to Claude.
2026-08-03Automated version bumping and added flexible installation options.3 changes
Feature
Added a global update flag and universal target
Users can now update the tool globally and install for a universal assistant standard.
Chore
Automated version bumping after releases
The release pipeline now updates internal version numbers automatically.
Chore
Bumped the tool to version 2.13.0
Released a new version number reflecting the latest changes.
2026-07-27A busy week expanding assistant compatibility and fixing installation and design-system behavior.6 changes
Feature
Added a Style Detail preview modal
The gallery gained a pop-up that previews UI styles with phone-style controls.
Fix
Improved Codex assistant installation
Skills now install correctly under Codex's shared folder structure.
Fix
Fixed design-system palettes and anti-patterns
Color palettes and warnings now correctly match each style's light or dark mode.
Fix
Generated Copilot slash-command files
The tool now produces the prompt file VS Code Copilot needs for its slash commands.
Fix
Ran the tool reliably without extra setup
Added a fallback so the tool runs through Node even when TypeScript isn't directly available.
Chore
Pinned the shadcn tool version
Locked a specific version of a design dependency for consistent results.
Dependencies and code review
Dependency advisories
Security Watch
No known advisories across 0 scanned dependencies.
No known advisories in the scanned dependencies.
Code review
No concrete code-level issues confirmed.
Get this report every week for your repos.
GitZoid learns each repo, reports what changed, and flags what needs attention. One flat price for the whole team.
$19 a month, flat · First 10 outputs free · No card required