agentleFS
mofirojeanmofirojean/angular-ui-skills/skills/ui-craft/SKILL.md

ui-craft

Skill33 starsChanged 3 months agoMITView on GitHub

Cross-cutting UI quality and dashboard-design skill. Covers the three tells of an amateur dashboard (data shapes the UI, progressive disclosure, hidden UI), plus the Refactoring UI principles (hierarchy, spacing, typography, color, depth, finishing touches) and how to apply each in Angular projects using Spartan/ng, PrimeNG, NG-ZORRO, or Angular Material. Use when the user asks to polish, refine, redesign, or "make this look better"; when building a dashboard, table, KPI surface, or data-heavy view; when the project has visual quality issues (low hierarchy, ambiguous spacing, grey-on-color text, color used as the only signal); or when the user mentions hierarchy, density, spacing, typography, palette, shadows, or empty states. Pairs with one of the library-specific skills (spartan-ng-developer, primeng-developer, ng-zorro-developer, angular-material-developer) for the implementation surface, and with angular-developer for the Angular fundamentals.

18 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use writing-for-agents, prototype and 3 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • MicrosoftSkill

    writing-for-agents

    Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.

    Also in 3 projects that use this

  • GoogleSkill

    prototype

    Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

    Also in 3 projects that use this

  • MicrosoftSkill

    handoff

    Compact the current conversation into a handoff document for another agent to pick up.

    Also in 4 projects that use this

  • GoogleSkill

    tdd

    Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

    Also in 3 projects that use this

  • +1 more from google

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 6 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 16 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Data drives the UI: Not the other way around. The shape, density, alignment, color, and chart-vs-table choice all come *from the data being …

    line 37
  2. Distinctive:

    When you're building or improving a dashboard, table, KPI surface, or data-heavy view, also read dashboard-patterns.md.

    line 56
  3. Distinctive:

    When to use a chart instead of a table (any time the data has a meaningful x-axis)

    line 62
  4. Distinctive:

    Critique before you generate: If the user has existing code, read it and call out the specific tells (grey-on-color label, no truncation, …

    line 70
  5. Very specific:

    Tailwind-first, design-token-aware: All four library skills compose with Tailwind.

    line 74
  6. Distinctive:

    Density is a choice, not a default: Default densities from component libraries are usually too loose for dashboards.

    line 76
  7. Very specific:

    No new components without reason: If a Helm/PrimeNG/NG-ZORRO/Material component covers it, use it.

    line 78
  8. Distinctive:

    Empty, loading, and error states are not optional: Every async surface gets all three.

    line 80

“Common” is measured: kept by at least 6 in 10 of the 16 edited versions of this file we could read. “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

UI Craft

Cross-cutting skill. Pair this with one of the library-specific skills (spartan-ng-developer, primeng-developer, ng-zorro-developer, angular-material-developer) plus the base angular-developer. This skill teaches what good looks like and why. The library skill teaches which component to reach for. Both load together when the user wants polished UI.

When this skill should fire

Activate eagerly when the user says any of:

  • "Make this look better / cleaner / more polished"
  • "This feels generic" / "looks like a template"
  • "Improve the dashboard" / "redesign this page"
  • "Why does this look amateur"
  • Specific topics: "hierarchy", "spacing", "typography", "color palette", "shadows", "empty state", "tooltip", "hover state", "loading state"

Also activate proactively (without being asked) when you see:

  • A dashboard or data-heavy surface being built
  • A table with all-grey text, no alignment, no muted/active distinction
  • Form labels in the same weight and color as the values
  • Color used as the only signal (no icon, no text, no shape backup)
  • Hover-only actions with no affordance hinting at them
  • Zero empty states / zero loading states for an async surface

The three tells of an amateur dashboard

These are the patterns that immediately give away a UI built without intent. Each gets its own reference file because each is a discipline, not a single rule.

  1. Data drives the UI. Not the other way around. The shape, density, alignment, color, and chart-vs-table choice all come from the data being displayed, not from a default table template. Read data-driven-ui.md.
  2. Progressive disclosure. Most actions, controls, and metadata don't belong permanently on screen. They belong tucked behind hover, popover, sheet, or context menu, with a clear affordance. The "spectrum of explicitness" runs from always-visible global actions to on-hover row actions to right-click-revealed power-user actions. Read progressive-disclosure.md.
  3. UI is what you can't see. Tooltips on every icon-only button, hover affordances on every clickable area, empty states, loading states, error states, the right-click menu, the keyboard shortcut, the toast on save. These hidden surfaces are most of the actual UI. Read hidden-ui.md.

The Refactoring UI principles

The book Refactoring UI by Adam Wathan and Steve Schoger gives the vocabulary for why a layout feels right. Each chapter maps to a reference here.

  • Hierarchy, Not all elements are equal. Emphasize by de-emphasizing. Labels are a last resort. Separate visual from document hierarchy. Read hierarchy.md.
  • Spacing, Start with too much whitespace, then use a sizing system. Don't fill the screen for the sake of filling it. Avoid ambiguous gaps. Read spacing.md.
  • Typography, Establish a type scale. Keep line-length in check (45-75ch). Line-height is proportional to font-size. Baseline-align, not center. Read typography.md.
  • Color, Ditch hex for HSL or OKLCH. Define a 9-step shade system per hue. Don't let lightness kill saturation. Greys don't have to be neutral. Color is never the only signal. Read color.md.
  • Depth, Emulate a light source. Shadows convey elevation. Two-part shadows (close + ambient). Flat designs can still have depth via overlap and value contrast. Read depth.md.
  • Finishing touches, Supercharge defaults. Add accent borders. Decorate empty backgrounds. Never skip empty states. Use fewer borders, more value contrast. Read finishing.md.

Dashboard-specific patterns

When you're building or improving a dashboard, table, KPI surface, or data-heavy view, also read dashboard-patterns.md. It covers:

  • KPI card patterns (label, value, delta, sparkline)
  • Table polish (chips for enumerated values, right-aligned numerics, truncated long text, muted inactive rows, sticky headers, hidden hover actions)
  • Filter bar conventions (toggle groups, comboboxes, date-range, save-as-view)
  • Sidebar nav density
  • When to use a chart instead of a table (any time the data has a meaningful x-axis)

Bridging to the library skills

The four library skills tell you which Helm / Aura / NG-ZORRO / Material component to use. This skill tells you how to make it look good. The bridge file maps each principle onto the four libraries, including which design tokens to override, which density mode to pick, and which common defaults are wrong out of the box. Read library-bridge.md.

Working principles

  1. Critique before you generate. If the user has existing code, read it and call out the specific tells (grey-on-color label, no truncation, no muted state) before refactoring. Naming the flaw teaches the user too.
  2. Show before/after when it matters. When you change spacing, hierarchy, or color, describe the why in one line so the change isn't mysterious. "Reviewers column trimmed from 152px to 96px so the avatars sit flush against the row's right padding instead of floating with empty space."
  3. Tailwind-first, design-token-aware. All four library skills compose with Tailwind. Use utility classes for one-off polish, but reach for theme tokens (text-muted-foreground, border-border, bg-card) so dark mode and theming come for free. Never hard-code hex.
  4. Density is a choice, not a default. Default densities from component libraries are usually too loose for dashboards. Tighten via the library's density token (Material: density-3, Spartan: smaller size variants, PrimeNG: --p-form-field-padding, ZORRO: nzSize="small"). See library-bridge.md.
  5. No new components without reason. If a Helm/PrimeNG/NG-ZORRO/Material component covers it, use it. Custom components ship with their own bugs.
  6. Empty, loading, and error states are not optional. Every async surface gets all three. Use the library's Skeleton / Empty / Alert primitives.
  7. Accessibility is part of craft. Focus rings, ARIA labels on icon-only buttons, keyboard shortcuts, contrast ratios. The accessibility.md in each library skill covers the mechanics; the intent lives here.

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.