Ethics review
Check a design for dark patterns and humane-design concerns. Use when the user asks whether a flow is manipulative, whether a pattern is a dark pattern, or for an ethics review on a sign-up, cancellation, paywall, notification, or persuasion-heavy surface.From its SKILL.md
npx -y skills add rafa-js/product-designer --skill ethics-reviewAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
3.4 KB, as published. Nobody here has run it
Ethics Review
Identify manipulative patterns and propose humane alternatives.
Inputs to confirm
- Surface: which flow / screen / pattern.
- Business intent: what is the team trying to optimize (signup, retention, ARPU, etc.)?
- Tradeoff appetite: is the team open to alternatives that lift retention but lower short-term conversion?
If the team is unwilling to consider tradeoffs, say so up front — an ethics review without willingness to act is theater.
Method
Walk these checks. For each finding, name the pattern, explain why it's harmful, and propose an ethical alternative.
1. Roach-motel test
- Is sign-up easier than cancellation?
- Is unsubscribing one click? Closing the account?
- Are subscription terms equally prominent in marketing and in the cancel flow?
2. Confirmshaming / coercive language
- Any opt-out worded to shame ("No, I don't want to save money")?
- Any default selection that nudges the user against their interest?
3. Urgency & scarcity
- Are countdowns real? (Do they reset on refresh?)
- Is "N people viewing now" based on data? Is "only N left" accurate?
- Are sale "was" prices true previous prices?
4. Defaults & opt-in
- Are data-sharing toggles opt-in (not opt-out)?
- Are upsells / extras pre-selected at checkout?
- Are notification permissions requested at the right moment with a real value prop, not on app launch?
5. Friction asymmetry
- Is friction added to actions that hurt the company (cancel, downgrade, export data) but removed from actions that help it (upgrade, share, invite)? Either parity or a deliberate decision in the user's favor.
6. Disclosure
- Is recurring billing disclosed before the free trial signup?
- Are auto-renewals announced before they charge?
- Are AI-generated outputs labeled?
- Are price changes communicated before they apply?
7. Engagement loops
- Streaks: do they reward use, or punish stopping? If breaking the streak feels like loss, redesign.
- Notifications: are they triggered by user-meaningful events or by inactivity?
- Infinite feeds: is there a natural stopping point or just decay?
8. Data and consent
- Is the minimum data collected, with the purpose explained at collection time?
- Can users export and delete their data?
- Are third-party trackers disclosed?
References: references/02-theory-and-principles/ethics-and-friction.md, references/04-ux-foundations/behavioral-economics.md (especially the dark/ethical comparison table).
Output
## Dark patterns found
| Pattern | Where | Why harmful | Ethical alternative |
## Borderline patterns
(persuasive but not unambiguously dark; describe tradeoffs)
## Strengths
(places the design is meaningfully humane — name these explicitly)
## Recommended changes, prioritized
1. ...
Principles to apply throughout
- The user's interest, the company's interest, and the law are three different bars. Name which you're applying.
- "It lifts conversion" is not a defense; many dark patterns lift conversion short-term and burn retention long-term.
- If you wouldn't be comfortable defending the choice publicly, don't ship it.