agentleFS
EndikaEndika/eskills/skills/ux-bar/SKILL.md

ux-bar

Skill0 starsChanged 57 days agoMITView on GitHub

Use when building or reviewing UI to apply my design and UX bar on top of frontend-design — semantic color, design-system rules (coral is brand; positive=owed, warn=owes, danger=error/delete), and responsive / horizontal-overflow checks.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. 0xabrar/gstack-distilled★ 14same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository14 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use brand sparingly — icon, hero, primary CTA — not to paint balances or errors.

    line 23
  2. Very specific:

    Semantics first: use the real element (button, a, nav, label) before a div + ARIA.

    line 37
  3. Distinctive:

    Keyboard: every interactive element reachable and operable by keyboard, with a visible focus indicator (SC 2.4.11).

    line 38
  4. Very specific:

    Errors in the interface language: never append an error's technical message after a translated prefix ("No se pudo guardar.

    line 41
  5. Very specific:

    When reviewing, report findings as file:line → which rule → the fix.

    line 81

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

ux-bar

Overview

My design/UX bar. It layers on top of frontend-design (reference it for the craft) and adds the rules I care about. Usable standalone on a diff, and invoked by the quality stage of eskills:task-flow.

Semantic color — the rule I break most often

  • Brand is brand, never status. Coral (#FF5A47 light / #FF7A66 dark) is the brand. Coral conventionally reads as "owe / alert" in money apps, so brand color must never also mean a debt or an error.
  • Map status to its own semantic token:
    • "te deben" / credit / owed-to-you → pos (green)
    • "tú debes" / owes → warn (amber)
    • error / delete / destructive / alert → danger (red)
  • Use brand sparingly — icon, hero, primary CTA — not to paint balances or errors.

Design-system discipline

  • Use tokens, never hardcoded palette values (no raw slate-/violet-/rose-NNN). Surfaces, text, border, brand all go through semantic CSS variables so light/dark flip for free.
  • Ship light and dark; verify both. Respect system as the default preference.

Accessibility floor (WCAG 2.2 AA)

POUR: Perceivable, Operable, Understandable, Robust. The checks that catch most barriers:

  • Contrast: small text 4.5:1, large text / UI components 3:1. A sub-AA choice is a finding unless named on purpose (e.g. coral fill + white text sits at ~3:1 — kept, with the darker #E8472F fallback documented).
  • Semantics first: use the real element (button, a, nav, label) before a div + ARIA. Correct Name / Role / Value; text alternatives for icons and images.
  • Keyboard: every interactive element reachable and operable by keyboard, with a visible focus indicator (SC 2.4.11). No keyboard traps; logical focus order.
  • Target size: interactive targets ≥ 24×24 CSS px (SC 2.5.8).
  • Dynamic state: announce async changes via aria-live/live regions; descriptive errors with how to fix (SC 3.3.3).
  • Errors in the interface language: never append an error's technical message after a translated prefix ("No se pudo guardar. Failed to save meeting"). Infrastructure returns a structured reason, the interface maps it to translated copy, and message stays for diagnostics only.
  • Reflow: usable up to 400% zoom without loss of content or function.

Responsive & horizontal overflow

The responsive bug I hit most. Symptom → cause → fix:

  • Grid/flex items don't shrink → the page overflows. A grid/flex item defaults to min-width: auto, so it refuses to go below its content's intrinsic min-size and drags the whole track wider than the viewport. Fix: min-width: 0 on the item (the grid/flex child) — not on the inner element.
  • Native controls keep an intrinsic width. <select>/<input> carry a min width (~longest option / ~20ch) that ignores min-width:0 on the control itself. Put min-width:0 on its grid/flex container; width:0 on the control alone does nothing.
  • Intrinsic grids beat breakpoints: repeat(auto-fit, minmax(min(BASIS, 100%), 1fr)). The min(BASIS,100%) is essential — it stops the column floor (BASIS) from overflowing below that width. Wraps N→1 with no magic media query.
  • overflow-x: clip is a mask, not a fix. It hides the overflow (and preserves position:sticky, unlike hidden), but the element is still too wide. Find and shrink the real culprit; keep clip only as defense-in-depth.
  • A header narrower than the content = overflow elsewhere. On mobile, horizontal overflow expands the layout viewport, so a max-width/centered header stops lining up with full-width cards. Don't fix the header — find what overflows.

Measure, don't eyeball. For any "not responsive" report: render at real widths and check documentElement.scrollWidth > clientWidth. Bisect the culprit by hiding each child (display:none) and re-measuring; confirm the fix by injecting the candidate style and re-measuring before editing. Verify down to 320px (real floor; Galaxy Fold cover ~280px) in every view. Harness: eskills:stack-gotchas → "mobile rendering in WSL".

UX checks

  • Destructive actions confirm or are undoable; loading/empty/error states exist.
  • Graceful degradation: a missing key/integration hides the feature, it doesn't crash.
  • Inclusive copy: UI strings follow the inclusive-language rule in eskills:standards — neutral/collective wording (no niños/as, no -x/-e), no assumed gender or family shape. Report offenders file:line → fix.

Output

When reviewing, report findings as file:line → which rule → the fix. Default to passing only what you actually checked.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.