ux-expert

Always-on UX advisor that surfaces relevant Laws of UX when building or modifying UI components. Proactively activates when creating, editing, or reviewing any user interface — components, layouts, navigation, forms, interactions, or visual design. Covers 30 laws across decision-making, cognition, visual organization, memory, engagement, and design principles.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "ux-expert" with this command: npx skills add shlomi-schwartz/shlomix/shlomi-schwartz-shlomix-ux-expert

UX Expert

Indexes 30 Laws of UX from lawsofux.com and proactively advises during UI/UX development. Searches the law index below, reads relevant law files, follows relationship links when warranted, and fetches external resources for deeper context.

When to Activate

Trigger this skill when:

  • Creating or modifying UI components, pages, or layouts
  • Adding options, choices, dropdowns, menus, or navigation items
  • Changing layouts, spacing, visual grouping, or element positioning
  • Modifying navigation flows, forms, or interactive elements
  • Reviewing or discussing UI/UX design decisions
  • Working on response time, loading states, or perceived performance
  • Handling touch targets, click areas, or element sizing
  • Any user-facing interface work where UX quality matters

Law Index

Decision & Choice

  • Choice Overload (choice-overload.md) — Users get overwhelmed with too many options; reduce, categorize, or add filtering
  • Hick's Law (hicks-law.md) — Decision time increases with number and complexity of choices; minimize options when speed matters
  • Tesler's Law (teslers-law.md) — Every system has irreducible complexity; absorb it in design so users don't have to
  • Postel's Law (postels-law.md) — Be liberal in what you accept from users, conservative in what you output

Cognitive Capacity

  • Cognitive Load (cognitive-load.md) — Total mental resources needed to use an interface; minimize extraneous load
  • Miller's Law (millers-law.md) — Working memory holds ~7 (plus/minus 2) items; chunk information into groups
  • Chunking (chunking.md) — Break information into meaningful groups to aid comprehension and recall
  • Working Memory (working-memory.md) — Temporary cognitive storage for active tasks; don't force users to remember across steps
  • Selective Attention (selective-attention.md) — Users focus on goal-relevant stimuli and miss the rest; design for what matters

Visual Organization (Gestalt)

  • Law of Proximity (law-of-proximity.md) — Near elements are perceived as grouped; use spacing to signal relationships
  • Law of Similarity (law-of-similarity.md) — Similar-looking elements are perceived as related; use consistent styling for groups
  • Law of Common Region (law-of-common-region.md) — Elements sharing a bounded area are perceived as grouped
  • Law of Uniform Connectedness (law-of-uniform-connectedness.md) — Visually connected elements are perceived as more related
  • Law of Pragnanz (law-of-pragnanz.md) — Users interpret complex visuals in the simplest form possible; favor clarity

Memory & Perception

  • Serial Position Effect (serial-position-effect.md) — Users best remember first and last items in a series; place key actions there
  • Von Restorff Effect (von-restorff-effect.md) — The distinct item among similar ones is most remembered; use for CTAs
  • Peak-End Rule (peak-end-rule.md) — Experiences are judged by their peak moment and ending, not the average
  • Zeigarnik Effect (zeigarnik-effect.md) — Incomplete tasks are remembered better; use progress indicators to drive completion
  • Cognitive Bias (cognitive-bias.md) — Systematic thinking errors that shape perception and decisions

Engagement & Motivation

  • Flow (flow.md) — State of deep immersion; balance challenge and skill, remove friction
  • Goal-Gradient Effect (goal-gradient-effect.md) — Effort increases as users approach a goal; show progress to motivate
  • Paradox of the Active User (paradox-of-the-active-user.md) — Users skip instructions and learn by doing; design for exploration
  • Parkinson's Law (parkinsons-law.md) — Tasks expand to fill available time; use constraints and deadlines

Design Principles

  • Jakob's Law (jakobs-law.md) — Users expect your site to work like others they know; leverage existing mental models
  • Aesthetic-Usability Effect (aesthetic-usability-effect.md) — Beautiful interfaces are perceived as more usable and forgive minor issues
  • Doherty Threshold (doherty-threshold.md) — Keep response times under 400ms to maintain user engagement and flow
  • Fitts's Law (fittss-law.md) — Target acquisition time depends on distance and size; make key elements large and close
  • Mental Model (mental-model.md) — Users' internal representation of how a system works; align design with expectations
  • Occam's Razor (occams-razor.md) — Prefer the simplest solution that meets requirements; remove unnecessary elements
  • Pareto Principle (pareto-principle.md) — ~80% of effects come from ~20% of causes; focus effort on high-impact elements

Slug notes: Fitts's Law = fittss-law.md, Hick's Law = hicks-law.md, Jakob's Law = jakobs-law.md, Miller's Law = millers-law.md, Tesler's Law = teslers-law.md, Postel's Law = postels-law.md, Law of Pragnanz = law-of-pragnanz.md

Search & Retrieval Process

  1. Extract context keywords from the current UI/UX work (component type, interaction pattern, design concern)
  2. Scan the law index above — match keywords, categories, and descriptions to the task
  3. Read 1-3 most relevant law files from references/laws/
  4. Check the ## Related section of each law read
  5. Decide whether to read related laws (see traversal rules below)
  6. When a law's ## Further Reading links are directly relevant, fetch them for deeper context (see deep dive rules below)
  7. Synthesize findings into actionable, specific advice

Relationship Traversal — When to Read Deeper

After reading a primary law, examine its ## Related section. Related laws use format [Law Name](/slug/) — map the slug to references/laws/{slug}.md.

Read a related law when:

  • The related law's topic directly applies to the current UI context
  • The primary law's takeaways reference concepts covered by the related law
  • The user's change touches multiple UX dimensions that span both laws

Skip a related law when:

  • It covers a UX dimension not relevant to the current task
  • The primary law already provides sufficient guidance
  • Reading more would delay actionable advice without adding value

Limits: Max 1 level of related-law traversal (never follow related-of-related). Max 4-5 total laws read per context (primary + related).

Deep Dive — When to Fetch Further Reading

Each law file has a ## Further Reading section with external links (Nielsen Norman Group, Smashing Magazine, etc.). These contain detailed research, examples, and case studies.

Fetch a Further Reading link when:

  • The user is making a significant design decision and needs evidence-based justification
  • The law's key takeaways alone are insufficient — the user needs implementation specifics or real-world examples
  • Multiple laws conflict or create tension and external context would help resolve the tradeoff
  • The user explicitly asks for deeper rationale or research backing

Skip fetching when:

  • The key takeaways in the law file are sufficient for the advice needed
  • The task is a minor UI tweak where general guidance is enough
  • Fetching would slow down time-sensitive advice without meaningful benefit

Limits: Max 2-3 external fetches per context. Prefer Nielsen Norman Group and Interaction Design Foundation links (highest signal-to-noise).

Output Format

  • Lead with the specific recommendation: what to do or what to change
  • Name the supporting law(s) — e.g., "Hick's Law suggests..."
  • Quote the most relevant key takeaway(s) from the law file
  • If multiple laws converge, note the pattern: "Both Hick's Law and Choice Overload point to..."
  • If a Further Reading source was fetched, briefly cite the key insight from it
  • Keep it concise: 3-6 sentences, not an essay
  • Only advise when a law is clearly relevant or being violated — never force-fit

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

Youtube Podcast Generator

Extracts the original text of Youtube video and converts it into a multi-voice AI podcast using a local Node.js API and FFmpeg. It also can show you the text...

Registry SourceRecently Updated
General

ERPClaw

AI-native ERP system with self-extending OS. Full accounting, invoicing, inventory, purchasing, tax, billing, HR, payroll, advanced accounting (ASC 606/842,...

Registry SourceRecently Updated
General

Whisper AI Audio to Text Transcriber

Turn raw transcripts into structured summaries, meeting minutes, and action items.

Registry SourceRecently Updated
General

Task Planner

- **name**: Task Planner. Use when you need task planner capabilities. Triggers on: task planner.

Registry SourceRecently Updated