Install
$ agentstack add skill-flitzrrr-frontend-design-skills-ux-design ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
UX Design
> Systematically designing user experiences -- from strategy to validated interaction.
Scope
Use this skill when: You are planning user research, creating personas, defining information architectures, documenting user flows, designing interaction patterns, or auditing existing products for UX quality.
Principles
1. The User Determines the Design -- Not the Designer
Every design decision is based on real user data. Assumptions without validation are speculation.
2. Design Experiences, Not Surfaces
UX encompasses the entire interaction -- from first perception to long-term usage.
3. Simplicity Beats Completeness
One feature that everyone understands is more valuable than ten features that nobody can find.
4. Emotions Are Measurable Success Factors
Positive experiences generate trust and return visits. UX is directly business-critical.
5. Iteration Over Perfection
Rapid prototyping, early testing, and continuous improvement beats extended planning phases.
Rules
The 5 Planes of User Experience (Garrett Model)
- DO: Work from bottom to top: Strategy > Scope > Structure > Skeleton > Surface.
- DO: At Plane 1 (Strategy), define: Who are the users? What are their goals? What are the business objectives?
- DO: At Plane 2 (Scope), prioritize features by user need, not by stakeholder wishes.
- DON'T: Jump directly to visual design (Plane 5) without defining the underlying planes.
User Research
- DO: Conduct at least 5 user interviews or usability tests BEFORE design.
- DO: Create data-driven personas: name, demographics, goals, pain points, context.
- DO: Use empathy maps as a quick format: What does the user think, feel, say, and do?
- DON'T: Create personas based on team assumptions. At least 3 real data points per persona.
- DON'T: Test only with colleagues. Always test with the actual target audience.
Information Architecture (IA)
- DO: Structure content according to the users' mental model, not the organizational structure.
- DO: Conduct card sorting to validate the content structure.
- DO: Keep navigation depth shallow: max. 3 clicks to reach target content.
- DO: Label navigation items in user language, not in industry jargon.
- DON'T: Use mega-menus with more than 7 main categories (Miller's Law: 7 +/- 2).
Interaction Design
- DO: Provide immediate feedback on every action (max. 400ms without feedback).
- DO: Make error handling helpful: WHAT went wrong and HOW to fix it.
- DO: Provide undo capability for destructive actions (min. 5-second undo window).
- DO: Reduce decisions per page (Hick's Law).
- DON'T: Show error messages only after form submission. Inline validation is the standard.
- DON'T: Use confirmation dialogs for trivial actions. Only for irreversible actions.
UX Laws
- DO: Fitts' Law -- Touch targets at least 44x44px, close to the expected location.
- DO: Jakob's Law -- Follow established conventions (users know other websites).
- DO: Doherty Threshold -- System responses under 400ms feel instantaneous.
- DO: Postel's Law -- Be liberal in what you accept, strict in what you produce.
User Flows
- DO: Define the primary flow (happy path) + at least 2 alternative paths.
- DO: Document flows as flowcharts with decision points.
- DO: Validate flows with real users before creating wireframes.
- DON'T: Create flows that lead to dead ends (pages without a next action or a way back).
Patterns
Detailed pattern descriptions: see references/patterns.md
Progressive Disclosure
Show only the essentials at the first level. Details behind "Learn more." Multi-step forms instead of one long form.
Onboarding Flow
Max. 3-5 steps. Progress indicator visible. Skip option. After completion: direct access to core value.
Search-Centric Design
Prominent search bar. Autocomplete with instant results. Filter options. Zero-results page with alternatives.
Feedback Loop
Visual feedback (hover, active), system feedback (loading indicator), result feedback (success message).
Empty States
Never leave blank areas. Always: explanation + call to action + appropriate visual.
Anti-Patterns
Feature Creep
Ever more features without validation. Solution: Every feature needs validated user demand. Under 20% usage = remove or hide.
Assumption-Driven Design
"I think the user wants..." Solution: At least 3 data points per key decision.
The Happy Path Only
Designing only for the ideal case. Solution: Define and design at least 3 error scenarios per flow.
Cognitive Overload
Too many options per page. Solution: Miller's Law (7 +/- 2). One page = one primary goal.
Ignoring Context
Same UX across all devices. Solution: Design context-dependently (mobile: on the go, one hand; desktop: more attention).
Checklist
- [ ] Strategy plane defined: user goals and business objectives documented?
- [ ] At least 3 data-driven personas created?
- [ ] User flows documented (happy path + edge cases)?
- [ ] Information architecture validated (card sorting or tree testing)?
- [ ] Navigation labels in user language (no internal jargon)?
- [ ] Navigation depth max. 3 levels?
- [ ] Immediate feedback on every action (under 400ms)?
- [ ] Error handling helpful (WHAT + HOW)?
- [ ] Undo for destructive actions?
- [ ] Inline validation on forms?
- [ ] Empty states with explanation and call to action?
- [ ] Touch targets at least 44x44px?
- [ ] Max. 7 main navigation items?
- [ ] No feature without validated user demand?
Cross-References
ui-design-- Visual implementation of UX conceptsaccessibility-- Inclusive design as a UX cornerstoneusability-- Testing methods and heuristicsnavigation-design-- IA implementation in navigation structurescustomer-journey-- UX in the overall context of the user journeydesign-process-- UX methods in the project workflowlanding-pages-- Conversion-optimized UX patterns
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: flitzrrr
- Source: flitzrrr/frontend-design-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.