[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7ASM2R5E46N4MA5114SM3RJ":3,"skill-same:01M7ASM2R5E46N4MA5114SM3RJ":48},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7ASM2R5E46N4MA5114SM3RJ","ui-ux-writing-forms-research-and-ethics",null,"Use when writing interface copy, designing forms and onboarding, planning or running UX research, auditing for dark patterns, or designing AI features. Covers UX-writing principles and the high-value surfaces, how localization affects design, form design and the checkout\u002Fconversion evidence, onboarding, choosing a research method, sample sizes, running a usability test that produces truth, metrics (SUS, HEART), the ethics line, the dark-pattern catalogue and the 2026 regulatory picture (DSA, FTC, DMA), and AI-era interfaces — design principles for AI features, generative UI and its risk, and tooling reality.","the-vibey-project\u002Fvibey","the-vibey-project","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F329298758?v=4","01M7AJHGZ9722CTJ4JNQZVRH0Q","the vibey project","01M7AJHGZVMG2K7TGXQ6XFHVG3",5,"MIT","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fui-ux-design-principles\u002Fskills\u002Fui-ux-writing-forms-research-and-ethics\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: ui-ux-writing-forms-research-and-ethics\ndescription: \"Use when writing interface copy, designing forms and onboarding, planning or running UX research, auditing for dark patterns, or designing AI features. Covers UX-writing principles and the high-value surfaces, how localization affects design, form design and the checkout\u002Fconversion evidence, onboarding, choosing a research method, sample sizes, running a usability test that produces truth, metrics (SUS, HEART), the ethics line, the dark-pattern catalogue and the 2026 regulatory picture (DSA, FTC, DMA), and AI-era interfaces — design principles for AI features, generative UI and its risk, and tooling reality.\"\n---\n\n# UI\u002FUX Design: UX Writing, Forms and Conversion, Research, Ethics, and AI-Era Interfaces\n\n> **Part 4 of 5** of the *UI\u002FUX Design Principles — Mobile, Tablet, Web, Desktop* reference (plugin `ui-ux-design-principles`), covering §10–§14. Sibling skills: `ui-ux-cognition-heuristics-and-navigation` (§0–§3), `ui-ux-interaction-layout-and-visual-design` (§4–§6), `ui-ux-design-systems-platforms-and-accessibility` (§7–§9), `ui-ux-reference` (§15–§20). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill.\n>\n> **Currency:** Verified August 2026. See §17 → `ui-ux-reference` for the currency snapshot and what goes stale first.\n>\n> **Related:** Overlaps the `frontend-design:graphic-ux-ui-design` skill. That one is oriented to implementation-time component and token decisions; this one is oriented to design reasoning, cross-form-factor principles, research method, and the regulatory landscape. Use both.\n\n> **How to read this.** Reference, not tutorial. Sections are independent. Three markers:\n> - **[DURABLE]** — grounded in perception, cognition, motor control, or decades of\n>   replicated research. Does not expire. Trust it.\n> - **[PLATFORM]** — a convention of Apple, Google, Microsoft, GNOME\u002FKDE, or the web.\n>   Verify against the current HIG; these change.\n> - **[CONTESTED]** — competent practitioners disagree, or the evidence is thinner than the\n>   confidence with which it's usually asserted. Both cases given.\n>\n> **⚠️ GOTCHA** boxes mark the mistakes that survive design review and get caught in\n> production, in court, or in the analytics.\n>\n> **The framing that organizes everything below:** an interface is a *proposal about how\n> a person should think*. Good design makes the proposal match how they already think.\n> Every principle in this document is a specialization of that idea — to a thumb, a\n> screen size, a screen reader, a legal regime, or an attention budget.\n\n---\n\n## §10. Content and UX Writing\n\n**[DURABLE] The interface is mostly words.** Most \"confusing UI\" is confusing *language*\nwearing a layout.\n\n### 10.1 Principles\n\n- **Clarity beats cleverness. Always.** Nobody has ever been delighted by a witty error\n  message they couldn't act on.\n- **Front-load the meaningful words.** Users scan the first two words of a link, heading,\n  or bullet. \"Download the annual report (PDF)\" not \"Click here to download…\"\n- **Use the user's vocabulary**, sourced from research, support tickets, and search logs —\n  not the database schema and not marketing.\n- **Be consistent.** One concept, one word, everywhere. If it's a \"workspace\" in nav it's\n  not a \"team\" in settings.\n- **Active voice, present tense, second person.** \"Your changes are saved\" not \"Changes\n  have been saved by the system.\"\n- **Sentence case** for UI labels reads faster than Title Case and is the modern convention\n  on every major platform.\n- **Numbers, dates, and units are localized.** `1,234.56` vs `1.234,56` is a real bug.\n\n### 10.2 The high-value surfaces\n\n**Buttons** — a verb naming the outcome. \"Save changes\", \"Create account\", \"Delete\nproject\". Never \"OK\", \"Submit\", \"Yes\u002FNo\" for consequential choices. The button label must\nmatch the heading that preceded it.\n\n**Error messages** — three parts, in this order:\n1. **What happened**, plainly. (\"We couldn't process your payment.\")\n2. **Why**, if you know and it's actionable. (\"Your card was declined by your bank.\")\n3. **What to do next.** (\"Try another card, or contact your bank.\")\nNever blame the user, never expose a stack trace or code as the only content (a support\nreference code *in addition* is fine and useful), never say \"invalid input\" without saying\nwhich input and what would be valid.\n\n**Empty states** — the most under-designed screen in most products, and a first-run user's\nfirst impression. It should: explain what belongs here, why it's empty, and give one clear\naction to fill it. Distinguish *first-use empty* (teach), *user-cleared empty* (celebrate or\nreassure), *no-results empty* (offer alternatives and a way to broaden), and *error empty*\n(diagnose).\n\n**Loading and progress** — say what's loading. \"Loading your invoices\" beats a spinner.\n\n**Confirmation and success** — name what happened and offer the next step. This is the\n\"end\" in peak-end (§1.4 → `ui-ux-cognition-heuristics-and-navigation`) and it's cheap to do well.\n\n**Microcopy on inputs** — a persistent hint below the field (not a placeholder) explaining\nformat requirements *before* the user fails, not after.\n\n### 10.3 Localization affects design, not just strings\n\n- **Text expansion**: German\u002FFinnish run 30–40% longer than English; some languages 100%+\n  for short strings. Fixed-width buttons break. Design for expansion or use flexible\n  layouts and test with pseudo-localization.\n- **RTL (Arabic, Hebrew, Farsi)** mirrors the entire layout — navigation, icons with\n  directionality (back arrows, progress), and reading order. Use *logical* properties\n  (`margin-inline-start`, not `margin-left`) and test with a force-RTL flag.\n- **Never concatenate translated fragments.** Word order differs; use positional format\n  strings.\n- **Plurals need real plural rules** (Arabic has six categories), not `if (n === 1)`.\n- **CJK** has different line-breaking, no spaces between words, and different comfortable\n  line-heights; ideographs need more vertical space.\n- **Names, addresses, phone formats, honorifics, and date orders are not universal.** A\n  required \"First name \u002F Last name\" split is wrong for a large fraction of the world.\n\n---\n\n## §11. Forms, Onboarding, and Conversion\n\n### 11.1 Form design — where usability turns directly into money\n\n**[DURABLE] Every field you remove increases completion.** Ask only for what you need *now*.\n\n| Rule | Why |\n|---|---|\n| **One column.** | Multi-column forms cause skipped fields and ambiguous tab order. Exception: genuinely paired short fields (city\u002Fstate, expiry month\u002Fyear). |\n| **Labels above fields, always visible.** | Fastest to scan; survives zoom; works with autofill; placeholder-as-label is a documented failure. |\n| **Group related fields** with clear spacing (§1.3 → `ui-ux-cognition-heuristics-and-navigation`). | Proximity does the work. |\n| **Inline validation *after* the field loses focus**, not on every keystroke. | Validating while typing tells users they're wrong before they've finished being right. |\n| **Show requirements up front** (password rules, format). | Error prevention > error handling (Nielsen #5). |\n| **Correct `autocomplete` attributes and `inputmode`.** | Autofill is a huge completion win and an accessibility feature. `inputmode=\"numeric\"` gives the right mobile keyboard. |\n| **Never mask what the user typed** (except passwords, and offer a reveal). | Card numbers, codes, and emails all benefit from being visible for checking. |\n| **Mark optional fields, not required ones** — if most are required. | Fewer asterisks, less noise. |\n| **Error summary at the top + inline errors + focus the first error.** | Screen-reader users and zoomed users can't see an error 2000 px down. |\n| **Preserve entered data on error.** | Losing a filled form is the fastest way to lose a user. |\n\n### 11.2 The checkout\u002Fconversion evidence\n\nBaymard Institute's long-running large-scale checkout usability research is the most-cited\ndata set here. Key figures, with the caveats they deserve:\n- **~70.2% average cart abandonment**, from a meta-analysis of ~50 studies (last updated\n  Sept 2025). This has been structurally stable for a decade, moving under a percentage\n  point in five years.\n- Baymard estimates the average large e-commerce site could gain **~35.26% conversion**\n  from checkout-design improvements alone — framed by them as an upper bound from a decade\n  of documented, solvable issues, **not a guaranteed return**, and explicitly excluding the\n  large share of abandonment that is pure browsing intent.\n- **Extra costs at checkout** (shipping, fees, tax) are consistently the top abandonment\n  reason (~39%); **forced account creation** ~24%; **too long\u002Fcomplicated a checkout**\n  ~19%; **not trusting the site with card details** ~19%.\n- The average checkout shows **~23.5 form elements \u002F ~14.9 fields**, versus an achievable\n  **12–14**; most sites can cut 20–60% of elements.\n\n**⚠️ Cite these carefully.** They are aggregates across many studies and verticals, mostly\nfrom one (excellent, commercial) research organization; live-behaviour trackers report\nhigher numbers (Dynamic Yield ~77.8% on a rolling 12-month basis) because they measure\nsomething slightly different. Use them to *prioritize*, not as targets your specific site\nwill hit.\n\n**The practical hierarchy:** show total cost early → offer guest checkout → cut fields →\none obvious next step per screen → visible trust signals near the payment field → support\nthe payment methods your market actually uses.\n\n### 11.3 Onboarding\n\n- **Show value before you ask for commitment.** Delay signup, delay permissions, delay the\n  tour. Permission requests should be *contextual* (\"Allow notifications so we can alert\n  you when your order ships\") and *at the moment of need*, not at first launch — a\n  cold-start permission prompt is the most reliable way to get a permanent denial.\n- **Progressive onboarding beats a carousel.** Teach the feature when it's first relevant.\n  Multi-screen intro carousels are almost universally skipped.\n- **Reduce time-to-first-value** relentlessly. Sample data, templates, and sensible defaults\n  beat an empty state plus a tutorial.\n- **Let users skip, and let them return.** A tour you can't exit is a hostage situation.\n\n---\n\n## §12. Research and Evaluation\n\n### 12.1 Choosing a method\n\n| Question | Method |\n|---|---|\n| Can people *do* the task? | **Usability test** (moderated or unmoderated) |\n| Why do they behave that way? | Interviews, contextual inquiry, diary study |\n| What do they do at scale? | Analytics, funnels, session replay |\n| Which version performs better? | A\u002FB test |\n| How do they think about the domain? | Card sort, tree test, mental-model interview |\n| Can they *find* it? | **Tree test** (IA without UI) \u002F first-click test |\n| Is it accessible? | Expert audit + AT testing + testing with disabled users (§9.5 → `ui-ux-design-systems-platforms-and-accessibility`) |\n| How do they feel over time? | Longitudinal survey (SUS\u002FUMUX-Lite\u002FNPS), diary |\n| Does it violate known principles? | **Heuristic evaluation** (§2.1 → `ui-ux-cognition-heuristics-and-navigation`) — cheap, fast, and finds different issues than testing |\n\n**[DURABLE] Attitudinal ≠ behavioural.** What people *say* they'd do is a poor predictor of\nwhat they *do*. Never ship a feature on survey preference alone.\n\n### 12.2 Sample sizes — the number everyone quotes and misuses\n\n**Nielsen & Landauer's \"5 users find ~85% of usability problems\"** is the most cited and\nmost abused finding in UX. What it actually says: with a problem-detection probability of\n~31% per user per problem, five users find about 85% of problems **in a single homogeneous\nuser group performing similar tasks**. What it does *not* say:\n- It doesn't apply across **distinct user segments** — each segment needs its own ~5.\n- It doesn't apply to **quantitative** measures (task time, success rate, satisfaction).\n  Those need 20+ per condition for anything resembling a confidence interval.\n- It doesn't mean five is *enough*; it means five is the point of diminishing returns for\n  *one round*, and **three rounds of five beats one round of fifteen** because you fix\n  things between rounds.\n- The 31% detection rate itself varies with task and interface complexity.\n\n**Rough guide:** qualitative usability 5–8 per segment per round; card sort 15–30;\ntree test 30–50; quantitative benchmark 20+ per condition; A\u002FB test — whatever your power\nanalysis says, calculated *before* you start.\n\n### 12.3 Running a usability test that produces truth\n\n- **Give tasks, not instructions.** \"Find out how much it would cost to ship two of these\n  to Berlin\" — not \"click the shipping calculator.\"\n- **Don't lead.** When they ask \"should I click here?\", answer \"what would you do if I\n  weren't here?\"\n- **Silence is data.** Let them struggle for a bit; the struggle is the finding.\n- **Watch behaviour, weight it above commentary.** Users are unreliable narrators of their\n  own difficulty and are systematically polite about your work.\n- **Recruit for the actual user**, not for whoever's convenient. Testing an enterprise tool\n  on your friends produces confident nonsense.\n- **Separate observation from interpretation** in your notes. \"Clicked Back three times\" is\n  an observation; \"was confused by the nav\" is a hypothesis.\n- **⚠️ Beware of testing only success paths.** Most real-world pain is in errors, edge cases,\n  recovery, and the second session — not the happy path you designed the prototype for.\n\n### 12.4 Metrics\n\n**HEART** (Google) — pick per-project, and pair each with a **Goal → Signal → Metric**:\nHappiness, Engagement, Adoption, Retention, Task success.\n\nComplementary standards:\n- **SUS** (System Usability Scale) — 10 items, 0–100; ~68 is average, 80+ is good.\n  Comparable across products and time, which is its real value.\n- **SEQ** (Single Ease Question) — one 7-point question after each task. Startlingly\n  informative for its cost.\n- **Task success rate, time on task, error rate** — the core behavioural triad.\n- **Core Web Vitals** (§6.6 → `ui-ux-interaction-layout-and-visual-design`) for web performance.\n\n**⚠️ Choose metrics that can go *down* when you make things worse.** Engagement metrics are\nnotorious for rewarding dark patterns: time-on-site rises when users are confused, and\nsession count rises when notifications are manipulative. Pair every engagement metric with a\nquality metric (task success, retention, complaint rate, unsubscribe rate).\n\n---\n\n## §13. Ethics, Persuasion, and Dark Patterns\n\n### 13.1 The line\n\n**[DURABLE] The test is whose interest the design decision serves when the user's and the\nbusiness's diverge.** Persuasion that helps a user do what they already wanted is design.\nDesign that exploits a cognitive bias to produce a choice the user would not otherwise make\nis a dark pattern — and increasingly, a legal violation.\n\n### 13.2 The catalogue\n\n| Pattern | Mechanism | Regulatory exposure |\n|---|---|---|\n| **Roach motel** (easy in, hard out) | Asymmetric effort | FTC: Epic $245M; Amazon $2.5B settlement (which surfaced internal emails on deliberately confusing cancellation) |\n| **Sneak into basket \u002F drip pricing** | Late disclosure | UCPD, FTC |\n| **Confirmshaming** (\"No thanks, I hate saving money\") | Social\u002Femotional pressure | DFA target |\n| **Preselected options \u002F pre-ticked boxes** | Default exploitation | GDPR consent invalidity |\n| **Asymmetric consent** (bright \"Accept all\", buried \"Reject\") | Choice architecture | **CNIL fined Google €150M and Microsoft €60M** on exactly this; TikTok €345M (Irish DPC) for public-by-default; Amazon €746M |\n| **False urgency \u002F fake scarcity** (\"1 room left!\") | Manufactured pressure | UCPD; DFA named target |\n| **Disguised ads** | Misrepresentation | FTC endorsement rules |\n| **Nagging** | Repetition until compliance | |\n| **Obstruction \u002F privacy zuckering** | Friction as a weapon | CPRA: \"dark patterns are about effect, not intent\" |\n| **Infinite scroll + autoplay + variable rewards** | Attention capture | DFA \"addictive design\" target |\n\n### 13.3 The regulatory picture (2026)\n\n- **EU DSA Article 25** already prohibits online-platform interfaces that \"deceive,\n  manipulate or otherwise materially distort\" users' ability to make free and informed\n  decisions. **DMA** carries fines up to 6% of global revenue for consent manipulation.\n- **EDPB Guidelines 03\u002F2022** define six GDPR dark-pattern categories: *overloading,\n  skipping, stirring, obstructing, fickle, left in the dark*. Useful as a design checklist,\n  not just a legal one.\n- **Consumer Rights Directive amendments** ban dark patterns in distance financial-services\n  interfaces — transposed by 19 December 2025, **applicable from 19 June 2026**.\n- **EU Digital Fairness Act** — confirmed in the Commission's 2030 Consumer Agenda (adopted\n  19 November 2025); **proposal expected late 2026**, targeting dark patterns, addictive\n  design, influencer marketing, and personalization. Adoption realistically 2027+, entry\n  into force 2028–2030. **Not law yet — do not describe it as such.**\n- **US FTC** is actively enforcing under Section 5, including **naming individual executives\n  as defendants**. Its 2024 study of 642 sites\u002Fapps found **76% used at least one possible\n  deceptive pattern** and ~67% used multiple. The Commission's 2022 EU study found **97% of\n  the most popular EU-used sites and apps** deployed at least one.\n- **California CPRA** treats dark patterns as consent-invalidating, with the CPPA's standard\n  being **symmetry**: the privacy-protective option must be *as easy* as the less protective\n  one.\n\n### 13.4 The design checklist\n\n- [ ] Cancelling is as easy as subscribing (same channel, same number of steps).\n- [ ] The privacy-protective choice is as prominent and as few clicks as the permissive one.\n- [ ] No pre-ticked consent boxes; consent is a positive, unambiguous act.\n- [ ] Total price, including all fees, disclosed before the user invests effort.\n- [ ] Urgency and scarcity claims are literally true and verifiable.\n- [ ] Decline options are neutrally worded — no shaming.\n- [ ] Defaults chosen for the user's benefit, and you can say why.\n- [ ] Ads and sponsored content are unmistakably labeled.\n- [ ] Consent flows are documented and evidenced.\n- [ ] No design element manipulates children or exploits known vulnerabilities.\n\n**[UNIVERSAL] \"It wasn't intentional\" is not a defense** — the CPPA's stated position is\nthat dark patterns are about *effect*, not intent. An A\u002FB test that increased conversion by\nmaking the reject button harder to find has produced a violation regardless of what anyone\nmeant.\n\n---\n\n## §14. AI-Era Interfaces\n\n### 14.1 What actually changed\n\n**[CONTESTED, and the most overclaimed area in the field.]** Two framings:\n- NN\u002Fg's line (Nielsen, 2023): generative AI is the **first new UI paradigm in 60 years** —\n  *intent-based outcome specification* rather than command specification. NN\u002Fg's **State of\n  UX 2026** extends this: UI is becoming less of a differentiator, AI-mediated interaction\n  sits on top of the interface, and equating UX with UI increasingly misdescribes the work.\n- The skeptical line: chat is a *regression* in discoverability, learnability, and\n  efficiency for most tasks; it externalizes the interface designer's job onto the user, who\n  must now guess what the system can do. A blank text box has zero affordances.\n\n**The synthesis practitioners actually ship:** a **hybrid** — conversational or intent-based\nentry, traditional UI for execution and refinement. The rule of thumb that circulates and\nholds up well: *if the user can describe the task in one sentence, conversational works; if\nthe task requires manipulation, comparison, precision, or spatial reasoning, give them a UI.*\n\n### 14.2 Design principles for AI features\n\n1. **Set expectations before the first interaction.** What can this do? What can't it?\n   What data does it see? Blank-box interfaces produce both over- and under-estimation.\n2. **Make capability discoverable** — suggested prompts, examples, templates. This is the\n   direct replacement for menus and buttons as an affordance mechanism.\n3. **Show provenance and uncertainty.** Citations, sources, confidence signals. Users\n   calibrate trust from these; without them they either over-trust or reject wholesale.\n4. **Keep the human in control**: editable output, regenerate, undo, and a way to *not* use\n   the AI path. Never make the AI the only route to a function.\n5. **Design for failure as a first-class state.** Hallucination, refusal, timeout,\n   irrelevance. These are not edge cases; they're the normal distribution's tail and users\n   hit them constantly.\n6. **Make cost and latency legible.** Streaming output is a UX decision (it converts a\n   10-second wait into a 300 ms first token, §6.6 → `ui-ux-interaction-layout-and-visual-design`). Token\u002Fcredit consumption should be\n   visible where it matters.\n7. **Feedback loops** — thumbs, corrections, \"not what I meant\" — both for model improvement\n   and because the ability to complain is itself a control affordance.\n8. **Privacy transparency** — what's sent, what's retained, what trains. This is a design\n   surface, not a legal footnote.\n\n### 14.3 Generative UI, and its risk\n\n\"Generative UI\" — interfaces assembled or adapted at runtime — is real and growing, and\nNN\u002Fg frames the shift as **outcome-oriented design**: designing adaptive *frameworks* that\nrespond to individual goals rather than optimizing single interfaces for an average user.\n\nTwo documented risks worth holding onto:\n- **Homogenization.** AI generation draws on dominant patterns; one Adobe study reported\n  **>42% of AI-generated interfaces shared similar navigation structures and components**.\n  The floor rises; the ceiling doesn't.\n- **Dark patterns at scale.** An NN\u002Fg survey found **36% of designers fear AI will\n  normalize dark patterns under the banner of UX optimization** — because an optimizer\n  pointed at conversion will find manipulation, and it will find it faster than a human.\n  If you let a system optimize choice architecture, constrain it with the §13.4 checklist\n  as hard constraints, not preferences.\n\n**Accessibility caution:** AI-generated UI and AI-generated code produce plausible markup\nwith unreliable semantics. Everything in §9 → `ui-ux-design-systems-platforms-and-accessibility` still has to be verified by a human. Figma's own\nframing of its AI outputs is that they still need human review for accessibility, semantics,\nand production readiness — take that at face value.\n\n### 14.4 Tooling reality (2026)\n\nDesign tooling has moved fast: Figma's **State of the Designer 2026** (survey of 906\ndesigners) reports **72% now use generative AI in their workflows**, with **98% having\nincreased usage in the past year** and **91% of those saying it improves output quality**,\nnot just speed. Figma's **Config 2026** introduced code layers on the canvas, design agents,\nand an MCP server that lets AI agents operate on the file with real design-system context;\n**Penpot** (open-source, self-hostable, native DTCG tokens, CSS-native output) shipped an\nMCP server too and is the credible option for regulated or open-source-constrained teams.\n\n**The honest read:** these tools compress production, not judgment. The bottleneck moved\nfrom making screens to knowing which screens to make — which is why the durable content in\n§1–§4 → `ui-ux-cognition-heuristics-and-navigation`, `ui-ux-interaction-layout-and-visual-design` and §12 is worth more now, not less.\n","# UI\u002FUX Design: UX Writing, Forms and Conversion, Research, Ethics, and AI-Era Interfaces\n\n> **Part 4 of 5** of the *UI\u002FUX Design Principles — Mobile, Tablet, Web, Desktop* reference (plugin `ui-ux-design-principles`), covering §10–§14. Sibling skills: `ui-ux-cognition-heuristics-and-navigation` (§0–§3), `ui-ux-interaction-layout-and-visual-design` (§4–§6), `ui-ux-design-systems-platforms-and-accessibility` (§7–§9), `ui-ux-reference` (§15–§20). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill.\n>\n> **Currency:** Verified August 2026. See §17 → `ui-ux-reference` for the currency snapshot and what goes stale first.\n>\n> **Related:** Overlaps the `frontend-design:graphic-ux-ui-design` skill. That one is oriented to implementation-time component and token decisions; this one is oriented to design reasoning, cross-form-factor principles, research method, and the regulatory landscape. Use both.\n\n> **How to read this.** Reference, not tutorial. Sections are independent. Three markers:\n> - **[DURABLE]** — grounded in perception, cognition, motor control, or decades of\n>   replicated research. Does not expire. Trust it.\n> - **[PLATFORM]** — a convention of Apple, Google, Microsoft, GNOME\u002FKDE, or the web.\n>   Verify against the current HIG; these change.\n> - **[CONTESTED]** — competent practitioners disagree, or the evidence is thinner than the\n>   confidence with which it's usually asserted. Both cases given.\n>\n> **⚠️ GOTCHA** boxes mark the mistakes that survive design review and get caught in\n> production, in court, or in the analytics.\n>\n> **The framing that organizes everything below:** an interface is a *proposal about how\n> a person should think*. Good design makes the proposal match how they already think.\n> Every principle in this document is a specialization of that idea — to a thumb, a\n> screen size, a screen reader, a legal regime, or an attention budget.\n\n---\n\n## §10. Content and UX Writing\n\n**[DURABLE] The interface is mostly words.** Most \"confusing UI\" is confusing *language*\nwearing a layout.\n\n### 10.1 Principles\n\n- **Clarity beats cleverness. Always.** Nobody has ever been delighted by a witty error\n  message they couldn't act on.\n- **Front-load the meaningful words.** Users scan the first two words of a link, heading,\n  or bullet. \"Download the annual report (PDF)\" not \"Click here to download…\"\n- **Use the user's vocabulary**, sourced from research, support tickets, and search logs —\n  not the database schema and not marketing.\n- **Be consistent.** One concept, one word, everywhere. If it's a \"workspace\" in nav it's\n  not a \"team\" in settings.\n- **Active voice, present tense, second person.** \"Your changes are saved\" not \"Changes\n  have been saved by the system.\"\n- **Sentence case** for UI labels reads faster than Title Case and is the modern convention\n  on every major platform.\n- **Numbers, dates, and units are localized.** `1,234.56` vs `1.234,56` is a real bug.\n\n### 10.2 The high-value surfaces\n\n**Buttons** — a verb naming the outcome. \"Save changes\", \"Create account\", \"Delete\nproject\". Never \"OK\", \"Submit\", \"Yes\u002FNo\" for consequential choices. The button label must\nmatch the heading that preceded it.\n\n**Error messages** — three parts, in this order:\n1. **What happened**, plainly. (\"We couldn't process your payment.\")\n2. **Why**, if you know and it's actionable. (\"Your card was declined by your bank.\")\n3. **What to do next.** (\"Try another card, or contact your bank.\")\nNever blame the user, never expose a stack trace or code as the only content (a support\nreference code *in addition* is fine and useful), never say \"invalid input\" without saying\nwhich input and what would be valid.\n\n**Empty states** — the most under-designed screen in most products, and a first-run user's\nfirst impression. It should: explain what belongs here, why it's empty, and give one clear\naction to fill it. Distinguish *first-use empty* (teach), *user-cleared empty* (celebrate or\nreassure), *no-results empty* (offer alternatives and a way to broaden), and *error empty*\n(diagnose).\n\n**Loading and progress** — say what's loading. \"Loading your invoices\" beats a spinner.\n\n**Confirmation and success** — name what happened and offer the next step. This is the\n\"end\" in peak-end (§1.4 → `ui-ux-cognition-heuristics-and-navigation`) and it's cheap to do well.\n\n**Microcopy on inputs** — a persistent hint below the field (not a placeholder) explaining\nformat requirements *before* the user fails, not after.\n\n### 10.3 Localization affects design, not just strings\n\n- **Text expansion**: German\u002FFinnish run 30–40% longer than English; some languages 100%+\n  for short strings. Fixed-width buttons break. Design for expansion or use flexible\n  layouts and test with pseudo-localization.\n- **RTL (Arabic, Hebrew, Farsi)** mirrors the entire layout — navigation, icons with\n  directionality (back arrows, progress), and reading order. Use *logical* properties\n  (`margin-inline-start`, not `margin-left`) and test with a force-RTL flag.\n- **Never concatenate translated fragments.** Word order differs; use positional format\n  strings.\n- **Plurals need real plural rules** (Arabic has six categories), not `if (n === 1)`.\n- **CJK** has different line-breaking, no spaces between words, and different comfortable\n  line-heights; ideographs need more vertical space.\n- **Names, addresses, phone formats, honorifics, and date orders are not universal.** A\n  required \"First name \u002F Last name\" split is wrong for a large fraction of the world.\n\n---\n\n## §11. Forms, Onboarding, and Conversion\n\n### 11.1 Form design — where usability turns directly into money\n\n**[DURABLE] Every field you remove increases completion.** Ask only for what you need *now*.\n\n| Rule | Why |\n|---|---|\n| **One column.** | Multi-column forms cause skipped fields and ambiguous tab order. Exception: genuinely paired short fields (city\u002Fstate, expiry month\u002Fyear). |\n| **Labels above fields, always visible.** | Fastest to scan; survives zoom; works with autofill; placeholder-as-label is a documented failure. |\n| **Group related fields** with clear spacing (§1.3 → `ui-ux-cognition-heuristics-and-navigation`). | Proximity does the work. |\n| **Inline validation *after* the field loses focus**, not on every keystroke. | Validating while typing tells users they're wrong before they've finished being right. |\n| **Show requirements up front** (password rules, format). | Error prevention > error handling (Nielsen #5). |\n| **Correct `autocomplete` attributes and `inputmode`.** | Autofill is a huge completion win and an accessibility feature. `inputmode=\"numeric\"` gives the right mobile keyboard. |\n| **Never mask what the user typed** (except passwords, and offer a reveal). | Card numbers, codes, and emails all benefit from being visible for checking. |\n| **Mark optional fields, not required ones** — if most are required. | Fewer asterisks, less noise. |\n| **Error summary at the top + inline errors + focus the first error.** | Screen-reader users and zoomed users can't see an error 2000 px down. |\n| **Preserve entered data on error.** | Losing a filled form is the fastest way to lose a user. |\n\n### 11.2 The checkout\u002Fconversion evidence\n\nBaymard Institute's long-running large-scale checkout usability research is the most-cited\ndata set here. Key figures, with the caveats they deserve:\n- **~70.2% average cart abandonment**, from a meta-analysis of ~50 studies (last updated\n  Sept 2025). This has been structurally stable for a decade, moving under a percentage\n  point in five years.\n- Baymard estimates the average large e-commerce site could gain **~35.26% conversion**\n  from checkout-design improvements alone — framed by them as an upper bound from a decade\n  of documented, solvable issues, **not a guaranteed return**, and explicitly excluding the\n  large share of abandonment that is pure browsing intent.\n- **Extra costs at checkout** (shipping, fees, tax) are consistently the top abandonment\n  reason (~39%); **forced account creation** ~24%; **too long\u002Fcomplicated a checkout**\n  ~19%; **not trusting the site with card details** ~19%.\n- The average checkout shows **~23.5 form elements \u002F ~14.9 fields**, versus an achievable\n  **12–14**; most sites can cut 20–60% of elements.\n\n**⚠️ Cite these carefully.** They are aggregates across many studies and verticals, mostly\nfrom one (excellent, commercial) research organization; live-behaviour trackers report\nhigher numbers (Dynamic Yield ~77.8% on a rolling 12-month basis) because they measure\nsomething slightly different. Use them to *prioritize*, not as targets your specific site\nwill hit.\n\n**The practical hierarchy:** show total cost early → offer guest checkout → cut fields →\none obvious next step per screen → visible trust signals near the payment field → support\nthe payment methods your market actually uses.\n\n### 11.3 Onboarding\n\n- **Show value before you ask for commitment.** Delay signup, delay permissions, delay the\n  tour. Permission requests should be *contextual* (\"Allow notifications so we can alert\n  you when your order ships\") and *at the moment of need*, not at first launch — a\n  cold-start permission prompt is the most reliable way to get a permanent denial.\n- **Progressive onboarding beats a carousel.** Teach the feature when it's first relevant.\n  Multi-screen intro carousels are almost universally skipped.\n- **Reduce time-to-first-value** relentlessly. Sample data, templates, and sensible defaults\n  beat an empty state plus a tutorial.\n- **Let users skip, and let them return.** A tour you can't exit is a hostage situation.\n\n---\n\n## §12. Research and Evaluation\n\n### 12.1 Choosing a method\n\n| Question | Method |\n|---|---|\n| Can people *do* the task? | **Usability test** (moderated or unmoderated) |\n| Why do they behave that way? | Interviews, contextual inquiry, diary study |\n| What do they do at scale? | Analytics, funnels, session replay |\n| Which version performs better? | A\u002FB test |\n| How do they think about the domain? | Card sort, tree test, mental-model interview |\n| Can they *find* it? | **Tree test** (IA without UI) \u002F first-click test |\n| Is it accessible? | Expert audit + AT testing + testing with disabled users (§9.5 → `ui-ux-design-systems-platforms-and-accessibility`) |\n| How do they feel over time? | Longitudinal survey (SUS\u002FUMUX-Lite\u002FNPS), diary |\n| Does it violate known principles? | **Heuristic evaluation** (§2.1 → `ui-ux-cognition-heuristics-and-navigation`) — cheap, fast, and finds different issues than testing |\n\n**[DURABLE] Attitudinal ≠ behavioural.** What people *say* they'd do is a poor predictor of\nwhat they *do*. Never ship a feature on survey preference alone.\n\n### 12.2 Sample sizes — the number everyone quotes and misuses\n\n**Nielsen & Landauer's \"5 users find ~85% of usability problems\"** is the most cited and\nmost abused finding in UX. What it actually says: with a problem-detection probability of\n~31% per user per problem, five users find about 85% of problems **in a single homogeneous\nuser group performing similar tasks**. What it does *not* say:\n- It doesn't apply across **distinct user segments** — each segment needs its own ~5.\n- It doesn't apply to **quantitative** measures (task time, success rate, satisfaction).\n  Those need 20+ per condition for anything resembling a confidence interval.\n- It doesn't mean five is *enough*; it means five is the point of diminishing returns for\n  *one round*, and **three rounds of five beats one round of fifteen** because you fix\n  things between rounds.\n- The 31% detection rate itself varies with task and interface complexity.\n\n**Rough guide:** qualitative usability 5–8 per segment per round; card sort 15–30;\ntree test 30–50; quantitative benchmark 20+ per condition; A\u002FB test — whatever your power\nanalysis says, calculated *before* you start.\n\n### 12.3 Running a usability test that produces truth\n\n- **Give tasks, not instructions.** \"Find out how much it would cost to ship two of these\n  to Berlin\" — not \"click the shipping calculator.\"\n- **Don't lead.** When they ask \"should I click here?\", answer \"what would you do if I\n  weren't here?\"\n- **Silence is data.** Let them struggle for a bit; the struggle is the finding.\n- **Watch behaviour, weight it above commentary.** Users are unreliable narrators of their\n  own difficulty and are systematically polite about your work.\n- **Recruit for the actual user**, not for whoever's convenient. Testing an enterprise tool\n  on your friends produces confident nonsense.\n- **Separate observation from interpretation** in your notes. \"Clicked Back three times\" is\n  an observation; \"was confused by the nav\" is a hypothesis.\n- **⚠️ Beware of testing only success paths.** Most real-world pain is in errors, edge cases,\n  recovery, and the second session — not the happy path you designed the prototype for.\n\n### 12.4 Metrics\n\n**HEART** (Google) — pick per-project, and pair each with a **Goal → Signal → Metric**:\nHappiness, Engagement, Adoption, Retention, Task success.\n\nComplementary standards:\n- **SUS** (System Usability Scale) — 10 items, 0–100; ~68 is average, 80+ is good.\n  Comparable across products and time, which is its real value.\n- **SEQ** (Single Ease Question) — one 7-point question after each task. Startlingly\n  informative for its cost.\n- **Task success rate, time on task, error rate** — the core behavioural triad.\n- **Core Web Vitals** (§6.6 → `ui-ux-interaction-layout-and-visual-design`) for web performance.\n\n**⚠️ Choose metrics that can go *down* when you make things worse.** Engagement metrics are\nnotorious for rewarding dark patterns: time-on-site rises when users are confused, and\nsession count rises when notifications are manipulative. Pair every engagement metric with a\nquality metric (task success, retention, complaint rate, unsubscribe rate).\n\n---\n\n## §13. Ethics, Persuasion, and Dark Patterns\n\n### 13.1 The line\n\n**[DURABLE] The test is whose interest the design decision serves when the user's and the\nbusiness's diverge.** Persuasion that helps a user do what they already wanted is design.\nDesign that exploits a cognitive bias to produce a choice the user would not otherwise make\nis a dark pattern — and increasingly, a legal violation.\n\n### 13.2 The catalogue\n\n| Pattern | Mechanism | Regulatory exposure |\n|---|---|---|\n| **Roach motel** (easy in, hard out) | Asymmetric effort | FTC: Epic $245M; Amazon $2.5B settlement (which surfaced internal emails on deliberately confusing cancellation) |\n| **Sneak into basket \u002F drip pricing** | Late disclosure | UCPD, FTC |\n| **Confirmshaming** (\"No thanks, I hate saving money\") | Social\u002Femotional pressure | DFA target |\n| **Preselected options \u002F pre-ticked boxes** | Default exploitation | GDPR consent invalidity |\n| **Asymmetric consent** (bright \"Accept all\", buried \"Reject\") | Choice architecture | **CNIL fined Google €150M and Microsoft €60M** on exactly this; TikTok €345M (Irish DPC) for public-by-default; Amazon €746M |\n| **False urgency \u002F fake scarcity** (\"1 room left!\") | Manufactured pressure | UCPD; DFA named target |\n| **Disguised ads** | Misrepresentation | FTC endorsement rules |\n| **Nagging** | Repetition until compliance | |\n| **Obstruction \u002F privacy zuckering** | Friction as a weapon | CPRA: \"dark patterns are about effect, not intent\" |\n| **Infinite scroll + autoplay + variable rewards** | Attention capture | DFA \"addictive design\" target |\n\n### 13.3 The regulatory picture (2026)\n\n- **EU DSA Article 25** already prohibits online-platform interfaces that \"deceive,\n  manipulate or otherwise materially distort\" users' ability to make free and informed\n  decisions. **DMA** carries fines up to 6% of global revenue for consent manipulation.\n- **EDPB Guidelines 03\u002F2022** define six GDPR dark-pattern categories: *overloading,\n  skipping, stirring, obstructing, fickle, left in the dark*. Useful as a design checklist,\n  not just a legal one.\n- **Consumer Rights Directive amendments** ban dark patterns in distance financial-services\n  interfaces — transposed by 19 December 2025, **applicable from 19 June 2026**.\n- **EU Digital Fairness Act** — confirmed in the Commission's 2030 Consumer Agenda (adopted\n  19 November 2025); **proposal expected late 2026**, targeting dark patterns, addictive\n  design, influencer marketing, and personalization. Adoption realistically 2027+, entry\n  into force 2028–2030. **Not law yet — do not describe it as such.**\n- **US FTC** is actively enforcing under Section 5, including **naming individual executives\n  as defendants**. Its 2024 study of 642 sites\u002Fapps found **76% used at least one possible\n  deceptive pattern** and ~67% used multiple. The Commission's 2022 EU study found **97% of\n  the most popular EU-used sites and apps** deployed at least one.\n- **California CPRA** treats dark patterns as consent-invalidating, with the CPPA's standard\n  being **symmetry**: the privacy-protective option must be *as easy* as the less protective\n  one.\n\n### 13.4 The design checklist\n\n- [ ] Cancelling is as easy as subscribing (same channel, same number of steps).\n- [ ] The privacy-protective choice is as prominent and as few clicks as the permissive one.\n- [ ] No pre-ticked consent boxes; consent is a positive, unambiguous act.\n- [ ] Total price, including all fees, disclosed before the user invests effort.\n- [ ] Urgency and scarcity claims are literally true and verifiable.\n- [ ] Decline options are neutrally worded — no shaming.\n- [ ] Defaults chosen for the user's benefit, and you can say why.\n- [ ] Ads and sponsored content are unmistakably labeled.\n- [ ] Consent flows are documented and evidenced.\n- [ ] No design element manipulates children or exploits known vulnerabilities.\n\n**[UNIVERSAL] \"It wasn't intentional\" is not a defense** — the CPPA's stated position is\nthat dark patterns are about *effect*, not intent. An A\u002FB test that increased conversion by\nmaking the reject button harder to find has produced a violation regardless of what anyone\nmeant.\n\n---\n\n## §14. AI-Era Interfaces\n\n### 14.1 What actually changed\n\n**[CONTESTED, and the most overclaimed area in the field.]** Two framings:\n- NN\u002Fg's line (Nielsen, 2023): generative AI is the **first new UI paradigm in 60 years** —\n  *intent-based outcome specification* rather than command specification. NN\u002Fg's **State of\n  UX 2026** extends this: UI is becoming less of a differentiator, AI-mediated interaction\n  sits on top of the interface, and equating UX with UI increasingly misdescribes the work.\n- The skeptical line: chat is a *regression* in discoverability, learnability, and\n  efficiency for most tasks; it externalizes the interface designer's job onto the user, who\n  must now guess what the system can do. A blank text box has zero affordances.\n\n**The synthesis practitioners actually ship:** a **hybrid** — conversational or intent-based\nentry, traditional UI for execution and refinement. The rule of thumb that circulates and\nholds up well: *if the user can describe the task in one sentence, conversational works; if\nthe task requires manipulation, comparison, precision, or spatial reasoning, give them a UI.*\n\n### 14.2 Design principles for AI features\n\n1. **Set expectations before the first interaction.** What can this do? What can't it?\n   What data does it see? Blank-box interfaces produce both over- and under-estimation.\n2. **Make capability discoverable** — suggested prompts, examples, templates. This is the\n   direct replacement for menus and buttons as an affordance mechanism.\n3. **Show provenance and uncertainty.** Citations, sources, confidence signals. Users\n   calibrate trust from these; without them they either over-trust or reject wholesale.\n4. **Keep the human in control**: editable output, regenerate, undo, and a way to *not* use\n   the AI path. Never make the AI the only route to a function.\n5. **Design for failure as a first-class state.** Hallucination, refusal, timeout,\n   irrelevance. These are not edge cases; they're the normal distribution's tail and users\n   hit them constantly.\n6. **Make cost and latency legible.** Streaming output is a UX decision (it converts a\n   10-second wait into a 300 ms first token, §6.6 → `ui-ux-interaction-layout-and-visual-design`). Token\u002Fcredit consumption should be\n   visible where it matters.\n7. **Feedback loops** — thumbs, corrections, \"not what I meant\" — both for model improvement\n   and because the ability to complain is itself a control affordance.\n8. **Privacy transparency** — what's sent, what's retained, what trains. This is a design\n   surface, not a legal footnote.\n\n### 14.3 Generative UI, and its risk\n\n\"Generative UI\" — interfaces assembled or adapted at runtime — is real and growing, and\nNN\u002Fg frames the shift as **outcome-oriented design**: designing adaptive *frameworks* that\nrespond to individual goals rather than optimizing single interfaces for an average user.\n\nTwo documented risks worth holding onto:\n- **Homogenization.** AI generation draws on dominant patterns; one Adobe study reported\n  **>42% of AI-generated interfaces shared similar navigation structures and components**.\n  The floor rises; the ceiling doesn't.\n- **Dark patterns at scale.** An NN\u002Fg survey found **36% of designers fear AI will\n  normalize dark patterns under the banner of UX optimization** — because an optimizer\n  pointed at conversion will find manipulation, and it will find it faster than a human.\n  If you let a system optimize choice architecture, constrain it with the §13.4 checklist\n  as hard constraints, not preferences.\n\n**Accessibility caution:** AI-generated UI and AI-generated code produce plausible markup\nwith unreliable semantics. Everything in §9 → `ui-ux-design-systems-platforms-and-accessibility` still has to be verified by a human. Figma's own\nframing of its AI outputs is that they still need human review for accessibility, semantics,\nand production readiness — take that at face value.\n\n### 14.4 Tooling reality (2026)\n\nDesign tooling has moved fast: Figma's **State of the Designer 2026** (survey of 906\ndesigners) reports **72% now use generative AI in their workflows**, with **98% having\nincreased usage in the past year** and **91% of those saying it improves output quality**,\nnot just speed. Figma's **Config 2026** introduced code layers on the canvas, design agents,\nand an MCP server that lets AI agents operate on the file with real design-system context;\n**Penpot** (open-source, self-hostable, native DTCG tokens, CSS-native output) shipped an\nMCP server too and is the credible option for regulated or open-source-constrained teams.\n\n**The honest read:** these tools compress production, not judgment. The bottleneck moved\nfrom making screens to knowing which screens to make — which is why the durable content in\n§1–§4 → `ui-ux-cognition-heuristics-and-navigation`, `ui-ux-interaction-layout-and-visual-design` and §12 is worth more now, not less.\n",{"name":5,"description":7},"develop","ae32736a1aabddabca5d2c96d665aa0ebc7d8700","src\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fui-ux-design-principles\u002Fskills\u002Fui-ux-writing-forms-research-and-ethics",3,"https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey","https:\u002F\u002Fthe-vibey-project.github.io\u002Fvibey\u002F","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project","https:\u002F\u002Fvibewithadam.matthewsteinberger.com","freedom for software engineers","2026-11-15 05:47:42",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add the-vibey-project\u002Fvibey --skill ui-ux-writing-forms-research-and-ethics",{"tool":40,"command":41},"bunx","bunx skills add the-vibey-project\u002Fvibey --skill ui-ux-writing-forms-research-and-ethics",{"tool":43,"command":44},"pnpm","pnpm dlx skills add the-vibey-project\u002Fvibey --skill ui-ux-writing-forms-research-and-ethics","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fui-ux-design-principles\u002Fskills\u002Fui-ux-writing-forms-research-and-ethics\u002FSKILL.md",true,[],[49,56,63,70,77,84],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7AT1FQJAE1FQYAMDTF490Y4","white-paper-writing","Use when writing white papers, technical reports, policy documents, thought leadership pieces, or any long-form authoritative document. Triggers on requests to write, structure, outline, or review white papers, technical reports, position papers, policy briefs, or evidence-driven arguments. Also use for B2B content marketing strategy, technical documentation requiring authority, academic-style reports, or any document that must educate while persuading. Use when someone asks \"how should I structure this\" for a long-form document.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fwhite-paper-writing\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7AT16BJRJJKK2MASXRZS76G","prose-technical-mechanics","Use when editing or writing prose at a technical\u002Fmechanical level: choosing between Anglo-Saxon and Latinate diction, placing information in topic vs stress positions, writing periodic vs cumulative sentences, handling sentence rhythm, punctuation as logic, transitions, paragraph structure, or phonesthetics. Use when someone wants to understand WHY prose works or doesn't, not just surface-level style advice. Triggers on word choice, sentence structure, emphasis, rhythm, diction, or any deep writing mechanics question.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fprose-technical-mechanics\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7AT0W9HT0ZZ9Y287NW99MR3","prose-mechanics","Use when writing, editing, or analyzing prose style; when trying to avoid generic AI-sounding writing; when applying specific prose styles such as journalistic, academic, narrative, KJV biblical, Hemingway plain, Latinate\u002FJamesian, Orwellian, Augustan, Victorian, Early Modern (Bacon, Browne, Milton), or free indirect discourse; or when coaching someone on writing voice and mechanics. Triggers on any request to improve, analyze, or write in a specific prose style, to fix \"AI-sounding\" writing, to make prose more vivid or concrete, to vary sentence rhythm, or to understand how a particular writer achieves their effects.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fprose-mechanics\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7AT0KR635JX66X2X6F3E2CP","narrative-structure","Use when structuring stories, pitches, presentations, white papers, or any persuasive narrative; analyzing the hero's journey in fiction or non-fiction; applying StoryBrand for business communication; designing narrative arcs for thought leadership; or understanding Vogler, Snyder, Propp, or Campbell frameworks. Triggers on any narrative structure, storytelling, pitch structure, StoryBrand, hero's journey, three-act structure, Freytag, Save the Cat, narrative arc, or \"how should I structure this story\u002Fpitch\u002Fpresentation\" request.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fnarrative-structure\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7AT0AKGW09NKKZDYBT8VRDE","legal-brief-writing","Use when writing legal briefs, motions, or any court filing; when analyzing judicial cognition and persuasion; when applying elite brief-writing techniques from Roberts, Clement, Kagan, Garner; or when structuring legal arguments for maximum persuasion. Also use for autistic\u002Fneurodivergent advocates who need to understand how their natural communication style maps to elite legal advocacy. Triggers on any legal writing, brief drafting, motion preparation, or legal argument structuring request.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Flegal-brief-writing\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7AT00VWPS4Q8B5VEMXX9NQT","journal-publishability-criteria","Use whenever a research paper, article or preprint is being written, revised, reviewed or published, and specifically before every publication of docs\u002Fpaper.md. Covers what legitimate journals actually gate on, in order of gating — the editor's desk screen (scope, novelty, package problems), what reviewers judge (methodological soundness, claims the evidence can carry, statistical and qualitative reporting), EQUATOR reporting guidelines, the ICMJE authorship criteria, COPE's view of text recycling, the near-universal generative-AI policy (no AI author, disclosure, author responsibility), TOP 2025 transparency standards, how to verify a journal is legitimate, submitting without an institution, and a pre-submission checklist. Triggers on requests to evaluate, review, prepare or submit a paper, to judge publishability, to write a declarations or data-availability section, to disclose AI use, or to check a journal.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fjournal-publishability-criteria\u002FSKILL.md",[],[]]