≈1k tokens on demand. Your agent reads this file only when SKILL.md points to it.
UX Anti-Patterns & Ethical Design
Purpose: Use this file when reviewing dark patterns, cognitive overload, accessibility violations, or ethical-design requirements.
Contents:
- dark-pattern catalog
- cognitive-overload catalog
- accessibility violation catalog
- ethical design principles and gates
Dark Patterns (DP-01 to DP-07)
| ID |
Pattern |
Harm |
Ethical alternative |
DP-01 |
Trick Questions |
accidental opt-in |
plain affirmative labels |
DP-02 |
Sneak into Basket |
unwanted charges |
explicit opt-in |
DP-03 |
Roach Motel |
hard cancellation |
symmetric signup and cancellation |
DP-04 |
Forced Continuity |
unnoticed billing |
reminders before renewal |
DP-05 |
Bait and Switch |
action/result mismatch |
transparent outcome disclosure |
DP-06 |
Privacy Zuckering |
excessive data sharing |
concise and accessible privacy choices |
DP-07 |
Confirm Shaming |
emotional manipulation |
neutral language |
Cognitive Overload (CO-01 to CO-06)
| ID |
Pattern |
Response |
CO-01 |
feature overload |
progressive disclosure |
CO-02 |
too many choices |
group and rank choices; 5-7 primary is a craft default, not a memory limit (_common/PROPORTION_AND_SPACING.md § Cognitive-capacity numbers) |
CO-03 |
visual noise |
content-first reduction |
CO-04 |
inconsistent UI patterns |
unify through the design system |
CO-05 |
dense information blocks |
hierarchy and whitespace |
CO-06 |
hidden navigation |
keep primary routes visible |
Accessibility Violations (AV-01 to AV-06)
| ID |
Violation |
Rule |
Response |
AV-01 |
low contrast |
WCAG 1.4.3, AA 4.5:1 |
verify text contrast |
AV-02 |
color-only status |
WCAG 1.4.1 |
add icon or text |
AV-03 |
hidden focus indicator |
WCAG 2.4.7 |
visible focus ring |
AV-04 |
unclear disabled state |
WCAG 1.4.11, AA 3:1 |
visible distinction plus semantics |
AV-05 |
keyboard-inaccessible interaction |
WCAG 2.1.1 |
full keyboard support |
AV-06 |
excessive motion |
WCAG 2.3.3 |
support prefers-reduced-motion |
Ethical Design Principles
- Respect user autonomy.
- Be transparent about data, AI, and consequences.
- Design inclusively with
WCAG 2.2 AA as the floor.
- Consider sustainability and unnecessary resource use.
- Favor long-term trust over short-term conversion tricks.
Quality Gates
- text contrast must meet
4.5:1 where required
- status must not rely on color alone
- cancellation should not be meaningfully harder than signup
- default settings should favor the user, especially for privacy
- motion-heavy proposals must specify reduced-motion behavior
Ethical AI Design Principles (2025-2026)
- Disclose AI involvement: users must know when content or decisions are AI-generated.
- Provide explainability: AI outputs must be explainable at a level appropriate to context.
- Embed fairness from the start (Ethics-by-Design): apply bias testing during data collection, not only post-deployment.
- Support human override: every AI-driven decision must have a clear path for human intervention.
- Ethical refusal is valid: declining to deploy GenAI can be the right decision.
AI-Specific Dark Patterns
| ID |
Name |
Impact |
Fix |
DP-08 |
AI Hallucination Concealment |
Trust erosion when discovered |
Disclose confidence levels; use Confidence Indicators |
DP-09 |
Invisible AI Agency |
Users unaware AI is acting on their behalf |
Surface AI actions via Proposal Cards; require explicit validation |
DP-10 |
Algorithmic Manipulation |
Personalization exploiting vulnerabilities |
Transparent recommendation logic; user-controllable personalization |
AI quality gates:
- AI-generated content must be labeled as such.
- Confidence indicators required for AI outputs in high-stakes contexts.
- Human override path must exist for every autonomous AI action.
- Bias testing must be documented before launching AI-driven personalization.
