Skip to content

jezweb/claude-skills / plugins/dev-tools/skills/ux-compare

ux-compare

Verified

Compare UX patterns across multiple reference apps using pattern libraries produced by ux-extract. Reads 2+ pattern-library.md files, walks them category by category, identifies where apps converge (strong signal), where they diverge (genuine design choice), what's unique to one app, and what's absent across the set. Produces an opinionated comparison document with recommendations for a new build. No browser needed — pure markdown analysis. Trigger with 'compare UX patterns', 'how do top apps handle X', 'ux comparison', 'pattern comparison across reference apps'.

FileKindSize
SKILL.mdskill md8.7 KB
references/comparison-template.mdmarkdown5.0 KB

references/comparison-template.md

View on GitHub
# Comparison Output Template

Structure for the document produced by `ux-compare`. Write to `docs/ux-comparisons/<topic>-YYYY-MM-DD.md`.

Topic slug matches the scope: `empty-states`, `onboarding`, `keyboard`, `destructive-actions`, `modals`, `whole-app`, etc.

```markdown
# UX Comparison: [Topic]

**Date**: YYYY-MM-DD
**Scope**: [What this comparison covers — specific feature, category, or whole-library]
**Compared**:
- [app-1]  (extracted YYYY-MM-DD)
- [app-2]  (extracted YYYY-MM-DD)
- [app-3]  (extracted YYYY-MM-DD — ⚠ stale, re-extract recommended)
**Source libraries**: `docs/ux-extracts/[app-1]/`, `docs/ux-extracts/[app-2]/`, etc.

## At a glance

[2–3 sentences: the headline findings. What converges? What diverges? What's surprisingly absent?]

## Convergence (strong signal)

Where most or all of the compared apps agree. These are defensible defaults — following them means users won't feel lost.

- **[Pattern]**: [how it's handled across the set, one sentence]
- **[Pattern]**: [...]

## Divergence (design decisions)

Where apps genuinely differ. Each divergence is a tradeoff — pick based on your context.

### [Pattern with divergence]

- **[App A]**: [what they do]
- **[App B]**: [what they do]
- **[App C]**: [what they do]

**Tradeoff**: [one sentence explaining when you'd pick each approach]

### [Another pattern]
[...]

## Unique approaches

Patterns only one app in the set uses. Call them out — either innovation worth borrowing, or weirdness worth avoiding.

- **[App]** — [pattern]. [Why they probably do it; whether it's borrowable]

## Absent across the set

What none or few of these apps do. Notable because it surfaces design *decisions by omission*.

- **[Absence]**: [X of N apps skip this]. [Interpretation — deliberate, convention, or overlooked?]

## Pattern-by-pattern detail

For each pattern category in scope, one section. Use this structure:

### [Pattern category, e.g. Empty list state]

**What each app does:**

- **[app-1]**: [concrete description — verbatim copy, layout, CTA behaviour]
  - *Source*: [path or anchor to the library entry, e.g. `ux-extracts/claude-ai/pattern-library.md#empty-states`]
- **[app-2]**: [...]
- **[app-3]**: [not covered in the extract — coverage gap]

**Convergent**: [what's shared across the set]

**Divergent**: [where the apps split]

**Unique**: [if any]

**Missing**: [if any app's library doesn't cover this]

### [Next pattern]
[...]

## Recommendations

This is the reason to have run the comparison. State positions. If you hedge here, the document is just a pile of bullets.

### For a new build in this space

1. **Safe default**: [the convergent choice — use this unless you have reason not to]
2. **Consider**: [divergent option A] — trades [benefit] for [cost]
3. **Consider**: [divergent option B] — trades [benefit] for [cost]
4. **Skip**: [unique approaches you'd avoid, with the reason]
5. **Fill a gap**: [absences worth addressing because your users would benefit]

### For an existing build in this space

- **Convergence you're missing**: [patterns common across references but not in your app]
- **Divergence worth reviewing**: [places your app does something unusual — is the reason still valid?]
- **Unique strengths**: [if your app already does something no reference does well, note it]

## Source library health

| Library | Extracted | Coverage for this comparison | Staleness risk |
|---------|-----------|----------------------------|----------------|
| [app-1] | YYYY-MM-DD | Full / Partial / Missing | Fresh |
| [app-2] | YYYY-MM-DD | Full | ⚠ 4 months — re-extract |

## What this comparison doesn't tell you

Be explicit about the limits:

- [e.g. "Performance was not captured in any extract — if speed is a key criterion, test directly."]
- [e.g. "Mobile patterns less thoroughly extracted for [app-3] — findings there may be partial."]
- [e.g. "All extracts used paid-tier accounts; free-tier UX may differ."]
```

## Guidelines for writing the comparison

- **Synthesise, don't list.** If you're just copying bullets from 5 libraries, you're not comparing — you're collating. The value is the *convergent / divergent / unique / absent* framing.
- **Be opinionated in recommendations.** *"It depends"* is the weakest possible output. State a position.
- **Cite sources.** Link to the specific section of the pattern library file. `ux-extracts/claude-ai/pattern-library.md#empty-states` is enough. Don't re-embed screenshots — the extract is the single source of truth.
- **Flag coverage gaps honestly.** "Not covered in app-3's extract" is not the same as "app-3 doesn't do this". One is a library gap, one is an app decision.
- **Date everything.** Both the comparison itself and each source library. Comparisons decay as fast as the libraries they're built on.
- **When app extracts are stale**, note it prominently — don't just mention it in the source table. Recommendations based on 6-month-old extracts should come with a qualifier.
- **Limit scope for sharpness.** A focused "empty states across 4 apps" is more useful than "whole-library comparison across 7 apps". Narrow wins.
brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).Stars: 322kUpdated Claude CodeCodexVerifiedInstall
brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).Stars: 198kUpdated Claude CodeCodexVerifiedInstall