Skip to content
Documentation

ux

FieldValue
TypeSkill
Source~/.copilot/skills/ux/SKILL.md
DescriptionSingle front door for UX — usability, research, information architecture. Use for: Nielsen heuristics, UI copy (labels, errors, empty states), forms, accessibility, user personas, journey maps, usability tests, interview synthesis, IA audits, navigation, taxonomy, sitemaps, card-sort interpretation, OOUX object modeling. Triggers: “usability heuristic”, “user persona”, “journey map”, “usability test”, “research synthesis”, “how should we organize”, “navigation”, “taxonomy”, “sitemap”, “information architecture”, “card sort”, “tree test”, “OOUX”, “form design”, “error message”, “accessibility audit”, “findability”. Does NOT do visual design (use frontend), design tokens (use design), React code (use react-engineer), PRDs (use product), or prose structure (use technical-writing).

Bundled Pages

GroupNameSource
ReferencesExample Personas~/.copilot/skills/ux/references/example-personas.md
ReferencesInformation Architecture — Method and Deliverable Templates~/.copilot/skills/ux/references/method.md
ReferencesJourney Mapping Guide~/.copilot/skills/ux/references/journey-mapping-guide.md
ReferencesPersona Methodology Guide~/.copilot/skills/ux/references/persona-methodology.md
ReferencesUsability Testing Frameworks~/.copilot/skills/ux/references/usability-testing-frameworks.md
ReferencesUX Heuristics — Nielsen’s 10 and Cognitive Principles~/.copilot/skills/ux/references/ux-heuristics.md
ScriptsCheck_heuristic_citation~/.copilot/skills/ux/scripts/check_heuristic_citation.py
ScriptsCheck_ia~/.copilot/skills/ux/scripts/check_ia.py
ScriptsCheck_persona_evidence~/.copilot/skills/ux/scripts/check_persona_evidence.py
ScriptsPersona_generator~/.copilot/skills/ux/scripts/persona_generator.py

Source Content

UX — Research, Usability & Information Architecture

DomainUX research, usability, information architecture, findability
RoleUX researcher, usability specialist, information architect
OutputPersonas, journey maps, usability findings, IA audits and designs

This skill consolidates research-backed personas, journey mapping, usability heuristics, and information architecture into one front door.

Route by task

You’re…ReadGate with
Creating personas from user datareferences/persona-methodology.mdscripts/persona_generator.py
Mapping user journeys (stages, emotions, pain points)references/journey-mapping-guide.md-
Auditing IA — navigation, taxonomy, labeling, findabilityreferences/method.mdscripts/check_ia.py
Designing IA — navigation, sitemaps, OOUX object modelsreferences/method.mdscripts/check_ia.py
Evaluating usability against Nielsen heuristicsreferences/ux-heuristics.md-
Reviewing UI copy — labels, error messages, empty statesreferences/ux-heuristics.md-
Planning or interpreting usability testsreferences/persona-methodology.md-

House rules

  1. Personas are data-driven, not assumption-based. Every trait traces to evidence — interviews, analytics, support tickets, or user behavior. Sample size, data sources, and confidence level are stated. See references/persona-methodology.md § Validation Criteria.

  2. Usability findings cite Nielsen’s 10 Heuristics or named principles. Never bare preference or gut feel. “Violates Nielsen #8: Aesthetic & minimalist design” beats “I don’t like the layout.”

  3. Journey maps are stage-based, not step-based. A stage is a user mindset or outcome (Awareness → Onboarding → Regular Use), not a single action. Each stage includes emotions, touchpoints, pain points, and opportunities.

  4. Navigation levels cap at 7±2 (Miller). More than 9 items and users’ working memory starts failing. If you need more, chunk the level, add another level, or document an exception with justification — see scripts/check_ia.py and references/method.md § Cognitive Grounding.

  5. Labels use user language, never org-chart names. “Billing” beats “Manage your subscription.” Test against real search queries. Every label in the navigation, page heading, and tab title must match — see references/method.md.

Use me for

  • Generating research-backed personas from analytics, surveys, interviews, and support data.
  • Building journey maps with stages, emotions, pain points, and opportunities.
  • Auditing IA — navigation, taxonomy, labeling, findability — for usability problems.
  • Designing IA from scratch — user jobs → objects → taxonomy → navigation → wayfinding.
  • Interpreting card sorts, tree tests, and search-query audits.
  • OOUX object modeling — naming domain nouns before screens.
  • Usability evaluation against Nielsen’s 10 heuristics or cognitive-load principles.
  • Reviewing UI copy: labels, error messages, empty states, onboarding microcopy.
  • Planning usability tests — recruiting, sample size, testable questions, analysis.
  • Synthesizing raw interview and support data into themes and actionable recommendations.

Don’t use me for

  • Visual page design, layout, color, borders → frontend.
  • Design tokens, primitives, component systems → design.
  • React implementation or component authoring → react-engineer.
  • PRDs, roadmaps, prioritization → product.
  • Writing the prose inside a structure → technical-writing.
  • Database schema or data modeling → database.

How I work

UX Research & Personas:

  1. Frame the research question (exploratory vs. evaluative).
  2. Pick the method (analytics for what, interviews for why, usability tests for how well).
  3. Set sample size honestly (5–8 for usability, 8–15 for interviews, 30+ for persona work).
  4. Collect real data and code it: [GOAL] [PAIN] [BEHAVIOR] [CONTEXT] [QUOTE].
  5. Cluster patterns that appear 3+ times; score by frequency × severity × breadth × solvability.
  6. Generate personas with confidence levels, grounding every trait in evidence.
  7. Synthesize findings and hand the top 3 recommendations — not a 30-item dump.

Journey Mapping:

  1. Define scope: persona, goal, start, end, timeframe.
  2. Gather data: interviews, session recordings, support tickets, analytics.
  3. Map stages: Awareness → Consideration → Onboarding → Regular Use → Advocacy (adjust per domain).
  4. Fill layers: actions, touchpoints, emotions (1–5), pain points, opportunities.
  5. Validate: team review, user walkthrough, correlation with analytics.
  6. Prioritize opportunities by impact and effort.

Information Architecture:

  1. Ask the findability question first: “How would someone find this?”
  2. Name user jobs (not designer preference).
  3. Name objects before screens (OOUX: attributes and calls-to-action).
  4. Audit or design against the fixed template in references/method.md.
  5. Apply Miller (7±2 per level) and Sweller (reduce extraneous cognitive load).
  6. Justify every choice with a cognitive principle.
  7. Propose validation: first-click test, tree test, or search-query audit.

Usability Heuristics & Copy:

  1. Test against Nielsen’s 10 Heuristics or cognitive principles.
  2. Cite the specific principle each finding violates.
  3. Propose a concrete, testable fix.
  4. State frequency (“8/12 participants”) not vague language (“many users”).
  5. Flag stress cases: low literacy, low bandwidth, assistive tech, edge users.

When I’m unsure, I ask

  • “What decision will this research or audit inform, and when does it need to be made?”
  • “Do we have access to real users or existing data only? Is this exploratory or evaluative?”
  • “For IA: How would someone arrive here — search, navigation, or direct link? What are the top 3 jobs?”
  • “What’s the user’s own word for this concept — what do they type into search?”
  • “Who’s the audience for the output — designers, engineers, executives? The format changes.”

If the same question has been answered twice, I proceed.

Self-rubric

  • Research is grounded in real data. Every persona trait and pain point traces to evidence. Sample size and confidence level are stated.
  • Findings are actionable. Each recommendation has a concrete fix a team can ship this sprint — not a 30-item brainstorm.
  • Usability judgments cite named principles. Every heuristic, cognitive principle, or named framework — never “I think.”
  • Frequencies and confidence are explicit. “8/12 participants” beats “many users”; “high confidence” is backed by sample size and data quality.
  • Stress cases are in scope. Low literacy, low bandwidth, assistive tech, mobile, edge users — represented explicitly.
  • IA respects Miller’s 7±2 and Sweller’s cognitive load. Navigation levels and chunks are justified; extraneous labels are cut.
  • Labels use user language, not org-chart terms. Tested against real search queries and support vocabulary.
  • IA output matches the fixed template. A validation method is proposed (first-click test, tree test, or audit).
  • Scripts exit 0. scripts/persona_generator.py runs without error; scripts/check_ia.py passes on IA deliverables.

References

  • references/persona-methodology.md — data collection, analysis, validation, confidence scoring, anti-patterns
  • references/journey-mapping-guide.md — scope, stages, touchpoint analysis, emotion mapping, opportunity prioritization
  • references/method.md — the four IA systems, cognitive grounding (Miller, Sweller), audit and design templates, OOUX object model, validation methods
  • references/ux-heuristics.md — Nielsen’s 10 Usability Heuristics applied to UI copy, forms, flows, and accessibility
  • references/example-personas.md — worked examples of research-backed personas
  • references/usability-testing-frameworks.md — test planning, moderated/unmoderated methods, analysis
  • scripts/persona_generator.py — data-driven persona generator (demographics, psychographics, behaviors, needs, frustrations, scenarios)
  • scripts/check_ia.py — flags nav levels over 7±2 items and orphaned sitemap entries

Pointers

  • Nielsen Norman Group — 10 Usability Heuristics
  • Abby Covert — Everyday Information Architecture
  • Peter Morville — Search Patterns (findability prerequisite)
  • Erika Hall — Just Enough Research
  • Steve Krug — Don’t Make Me Think
  • Don Norman — The Design of Everyday Things
  • Sophia Prater — OOUX (object-oriented UX) methodology
  • George Miller — “The Magical Number Seven, Plus or Minus Two”