UI UX Pro Max Skill Review: Verdict, Safety and Alternatives

The UI UX Pro Max skill (ui-ux-pro-max-skill) pairs design rules with a local Python search over design catalogs. Our verdict is Conditional, 68/100, with safety grade B.

Get the skill: ui-ux-pro-max folder on GitHub · SKILL.md at the reviewed commit · skills.sh listing

What is the UI UX Pro Max skill?

The UI UX Pro Max skill is nextlevelbuilder's ui-ux-pro-max-skill. It pairs a priority table of design rules with a local Python search. The search covers catalogs of styles, palettes, font pairings and UX guidelines.

An Agent Skill is a folder of instructions that an agent loads when a task matches its description. You can read what a SKILL.md contains before going further. UI UX Pro Max is a skill with scripts attached. Its SKILL.md holds the rules, and three of its five Python scripts run the search.

The skill's job is to give a coding agent concrete design decisions before it builds an interface. Its users are developers who build web, mobile or desktop UI without a designer. They work in Claude Code, Codex, Cursor and similar agents.

The rule table ranks categories by priority. Accessibility comes first, with contrast, touch targets and focus states. The search then supplies the specific tokens: hex palettes, font pairings, spacing scales and anti-patterns.

Which UI UX Pro Max sites are official?

The official UI UX Pro Max sites are the GitHub repo and uupm.cc, which the README links. The publisher's uupm.cc sells a Premium tier on top. The site ui-ux-pro-max-skill.com calls itself unofficial.

Three sites carry the skill's name. Only two of them come from the publisher.

Four directories mirror the README or SKILL.md: ui-skills.com, mdskills.ai, uichemy and mcpmarket. None of them is the publisher.

How does UI UX Pro Max search its design data?

UI UX Pro Max runs search.py, a BM25-style search over bundled CSV files. It matches your brief to a style, palette, font pairing and landing pattern. The agent then builds from that design system.

BM25 is a keyword ranking method from search engines. Here it ranks rows in local CSV files, with no model call. The flow has four steps:

1. Describe the product, industry and stack in your request.
2. Let search.py, core.py and design_system.py rank the matching CSV rows.
3. Read the returned design system: hex tokens, fonts, anti-patterns and a landing structure.
4. Add --persist to write design-system/<slug>/MASTER.md for later sessions.

The catalog counts differ across the sites that describe the skill. The v2.15.0 README figures are the publisher's own, and we did not recount them.

Is the UI UX Pro Max skill worth installing?

Yes for new pages, but test it before a redesign. The UI UX Pro Max skill scores 68/100, Conditional, at Moderate confidence. Tessl measured a 1.86x uplift, but issue #263 shows cluttered pages hiding main content.

Tessl is a skills registry that publishes evals and quality scores. In its two-scenario eval, this large, well-maintained design database lifted scores 1.86x. Field reports show cluttered or regressed pages, and a script path that resolves only under plugin installs.

Its main strength is turning vague briefs into specific tokens. The skill returns hex palettes with on-colour pairs, font pairings, spacing scales and anti-patterns. Its main weakness is output that turns over-decorated and buries main content. Issue #263 is accepted and still unfixed.

It fits developers starting a new page or product. They get a concrete palette, font pairing, style and accessibility checklist picked for their industry. It is a poor fit in three cases:

• Existing, well-designed sites, where users report a heavier template look after polishing.
• Teams that want distinctive art direction over industry-typical defaults.
• Environments without Python, or where every shell command needs approval.

Confidence is Moderate. The missing piece is a live A/B on current models. It needs to cover redesigns, dashboards and non-plugin installs, and check whether search.py runs.

Our page on how verdict scores are set has the weights. The best UI/UX design skills, ranked puts this verdict beside its peers.

How many tokens does UI UX Pro Max use?

The UI UX Pro Max file measures about 3,992 tokens per load. Each search prints an estimated 1,000 to 3,000 more. Our estimate for a typical task is 12,000 input and 6,000 output tokens.

The typical task is a design system plus two or three searches for a new landing page. We estimate 5 such tasks a week.

The hidden cost is approval time. Each search call can trigger a permission prompt in Claude Code. The 3.6 MB folder carries tests and catalogs that #484 found unused.

What changed since our review?

The UI UX Pro Max skill file has not changed since our review. No release followed v2.15.0, and npm is still stuck. Two fixes changed Cursor and Copilot installs, so the verdict stands.

We checked the repo through the GitHub API and the npm registry on 2026-10-21:

• 2026-10-03: commit 477bcb2 (PR #509) stopped the Cursor skill pointing at an empty Quick Reference.
• 2026-10-12: commit 7246805 (#527) added OpenClaw platform support.
• 2026-10-15: commit 56ea9cf (#533) made Copilot install the skill in .github/skills.
• 2026-10-21: the latest release is still v2.15.0, and npm ui-ux-pro-max-cli is still 2.15.0.
• 2026-10-21: the core SKILL.md hash is unchanged (ea087c34…), and stars reached 137,944.

Fixes #523, #526 and #532 touched sibling design, brand and icon skills, not this one. Issues #263, #446 and #457 remain open.

Is the UI UX Pro Max skill safe to install?

Yes, with care: the UI UX Pro Max skill carries safety grade B. Its five Python scripts search bundled CSVs locally, without network calls or credentials. Files are written only with --persist.

Grade B means the skill ships scripts, and nothing in them looked alarming. Our scan of the pinned commit found these facts:

• 73 files, about 3.6 MB: SKILL.md, 2 reference docs, CSV and JSON data, 5 scripts and 13 test files.
• search.py, core.py and design_system.py run BM25-style search with the Python standard library only.
• No HTTP requests happen at runtime. urllib.parse appears only for URL checks in validate_data.py and tests.
• No credentials are needed. The scripts read only COLORTERM and the plugin-root variable.
• Writes happen only with --persist, using exclusive create, so existing files stay unless you add --force.
• SKILL.md tells the agent to treat search results as recommendations and to keep private project data out of queries.
• Commands use ${CLAUDE_PLUGIN_ROOT} paths, which stay empty outside a plugin install.
• The test suite and validator ship inside the installed skill as dev tooling.
• The skills.sh audits (Gen Agent Trust Hub, Socket, Snyk) all pass, and Tessl's Snyk scan passed.

One directory, mdskills.ai, lists Network Access as a required permission. Our scan found no runtime network call, so we treat that listing as wrong. Issue #333 reports Claude flagging an installed settings.local.json as malicious. Maintainers say they fixed it earlier. Our guide shows how to vet a skill's scripts before installing.

What does the community report about UI UX Pro Max?

Community reports on UI UX Pro Max are split. Tessl measured 1.86x on Sonnet 4.6, and skiln.co was favourable in three tests. A note.com test preferred Frontend Design in two of three industries.

These dated reports come from GitHub, Hacker News and the web. Strong items come first.

What are UI UX Pro Max's known failures?

At v2.15.0, UI UX Pro Max has four known failures. It can bury main content in busy layouts (#263) and worsen existing sites (#446). It fires on half of relevant prompts, and its script path breaks outside plugins.

The worst failure for this job is a cluttered design system. The agent then ships a page worse than the one it replaced. We found four failures and one distribution bug:

1. Busy layouts. Issue #263 shows pages that turned chaotic and hid the main information. Maintainers accepted it, and it is still open.
2. Regressed redesigns. Issue #446 reports that applying the skill to an existing site made the design worse.
3. Missed triggers. Issue #484 measured 50% recall with 0 false positives in 120 negatives. The agent never opened the reference files.
4. Broken script paths. SKILL.md calls ${CLAUDE_PLUGIN_ROOT}, which is empty outside plugin installs. The agent then builds without its data.
5. Stuck npm releases. Issue #457 reports nothing published to npm since 2.15.0.

In #484, the chart case lost to the dataviz skill in 20 of 20 attempts. Our guide covers what to do when a skill does not trigger.

Which agents does UI UX Pro Max work with?

UI UX Pro Max is verified only in Claude Code, on Sonnet 4.6, Opus 4.7 and Haiku 4.5. The publisher claims 15 agents, including Cursor, Codex and Copilot. Every agent needs Python 3 and a shell.

Verified means our evidence shows the skill running there. Claimed means only the publisher says so.

The publisher's site names 8 agents, and mdskills lists 16. We list the 15 agents the publisher claims. Our hub lists Claude Code skills by verdict. The Claude apps run skills differently, as our page on skills in the Claude desktop and web apps explains.

How do you install the UI UX Pro Max skill?

Install the UI UX Pro Max skill in Claude Code with two plugin commands. Add the nextlevelbuilder marketplace, then install ui-ux-pro-max. The plugin route is the one its script paths expect.

Check that Python 3 runs in your shell first. Every route below depends on it.

In Claude Code

Run these two commands inside Claude Code:

/plugin marketplace add nextlevelbuilder/ui-ux-pro-max-skill
/plugin install ui-ux-pro-max@ui-ux-pro-max-skill

The plugin install sets ${CLAUDE_PLUGIN_ROOT}, so the script paths in SKILL.md resolve. Default permissions refuse the search calls: #484 saw 83 of 83 calls refused. Approve the Python calls, or add a permission rule for search.py. Our full Claude Code skills setup covers permission rules.

In Cursor, Codex, Antigravity and other agents

The publisher's CLI installs the skill for other agents:

npm install -g ui-ux-pro-max-cli
uipro init --ai cursor

Swap cursor for codex or another agent name. The npm package is still 2.15.0 because publishing has failed since August (#457). The skills CLI is a second route, listed on ui-skills.com:

npx skills add https://github.com/nextlevelbuilder/ui-ux-pro-max-skill --skill ui-ux-pro-max

Two post-review commits changed these installs. PR #509 fixed the Cursor template, and #533 moved Copilot installs to .github/skills. Our evidence does not cover these routes in these agents. Confirm that search.py runs after install, since the plugin-root path is empty there.

How do you use the UI UX Pro Max skill?

Use the UI UX Pro Max skill by naming the product, industry and stack in your request. The agent runs the search, returns a design system, then builds. Confirm that search.py ran.

The README describes two modes. Skill mode fires on its own when a UI task matches. Workflow mode starts from a slash command. The steps are the same:

1. Name the product, industry and stack, such as "fintech dashboard in Next.js".
2. Let the agent run search.py with --design-system.
3. Review the palette, fonts and anti-patterns before any code.
4. Build the page from that design system.
5. Check that the main content stays above the fold.

Three tasks have evidence behind them:

• Design system for a new page (primary, weekly). Tessl's "DataPulse Landing Page" scenario covers it.
• Accessibility remediation of forms (primary, occasional). Tessl ran a registration-form scenario.
• Chart choice for dashboards (secondary). The dataviz skill won this case 20 of 20 times in #484.

Test it yourself before your team adopts it:

• Install it the way your team will.
• Ask for one landing page.
• Ask for one redesign of an existing page.
• Check whether search.py ran.
• Check whether main content stays above the fold.

What are the alternatives to UI UX Pro Max?

The alternatives to UI UX Pro Max are Emil Design Engineering, Frontend Design, Impeccable and Taste Skill. The first two score 76, Recommended; Impeccable (67) and Taste Skill (62) are Conditional. Frontend Design gives one visual direction, not a database.

The table is sorted by score. Each row names the case where that skill does better.

Our head-to-head page, UI/UX Pro Max vs Frontend Design, covers the two side by side.

Directory ratings differ from ours. mdskills.ai shows an 8.0/10 advisor score with no stated method. Our score comes from dated evidence and a fixed rubric. The design hub lists every design skill we rated.

Where is UI UX Pro Max on GitHub?

UI UX Pro Max lives in the nextlevelbuilder/ui-ux-pro-max-skill GitHub repo under .claude/skills/ui-ux-pro-max. It is MIT-licensed, with 73 files in about 3.6 MB. We reviewed release v2.15.0 at commit 09170ee on 2026-09-30.

• Repo: github.com/nextlevelbuilder/ui-ux-pro-max-skill, subpath .claude/skills/ui-ux-pro-max.
• Licence: MIT.
• Releases: 41 in 12 months and 14 in 90 days. The latest is v2.15.0 (2026-08-13).
• Contributors: 92.
• Stars: 131,959 at review and 137,944 on 2026-10-21.
• Packages: a Claude Code plugin manifest and the npm package ui-ux-pro-max-cli.
• Installs: 377.4K on skills.sh at review.

Maintainability scores 80. Maintainers merge fixes quickly, as #487, #500 and #506 show. A broken npm pipeline and a bulk close of outside PRs on 2026-08-12 weigh against that.

Common questions

Is UI UX Pro Max free?

Yes. The repo is MIT-licensed, and uupm.cc sells a separate Premium tier for brand, logo, banner and slide work.

Does UI UX Pro Max need Python?

Yes, the search needs Python 3 and a shell. Without them, the skill falls back to its static rule table.

Does the UI UX Pro Max search send my project data anywhere?

No. Our scan found no runtime network calls. SKILL.md tells the agent to keep private project data out of queries.

Does UI UX Pro Max work in ChatGPT?

No. ChatGPT is not on the publisher's list, and the skill needs a local shell to run search.py.

Is ui-ux-pro-max-skill.com the official site?

No. Its header calls it an unofficial translation site, while the publisher's site is uupm.cc.