<!--
  Landing Page Pack
  From TemplatedAI.io — all the prompts the gurus sell, rebuilt in plain
  language and honestly flagged. 6 skills, 0 tested. Empty folder to a live, fast, accessible landing page — bones, skin, sections, form, performance pass. No framework, no template-shop markup.

  HOW TO USE: paste this whole file into a Claude Project, ChatGPT Project,
  Gemini Gem, or Cursor as knowledge/instructions. Then ask your AI to use
  a skill by name.

  HONESTY: "tested: no" in a skill means nobody has personally run it and
  pasted a real result yet. Use it — just judge the output yourself.
-->

---
name: Semantic HTML Skeleton
category: coding
works_with: [claude, chatgpt, gemini, cursor]
difficulty: plug-and-play
version: 1.0
source: adapted (viral "web architect" prompt pack, rewritten; guru naming and puffery removed)
tested: no
---

# Semantic HTML Skeleton

## What it does
Generates a clean, accessible HTML5 page structure built from real semantic tags (header, nav, main, section, article, footer) instead of a pile of divs — so the page is readable, screen-reader friendly, and doesn't jump around while it loads.

## When to use it
- Starting a new page and you want the bones right before any styling
- Inheriting a div-soup layout you want to restructure
- You care about accessibility and Core Web Vitals, not just looks

## The skill
```
Act as a front-end developer who writes semantic, accessible HTML5.
Rules: use semantic elements (header, nav, main, section, article,
aside, footer) and reserve <div> only for genuine styling wrappers
with no semantic meaning; every image gets alt text; headings nest
in order (one <h1>, then h2s, etc.); interactive elements are real
buttons/links, not clickable divs. Don't inline styles or scripts.
No claims about speed you can't back — just clean structure.

What the page is for: [DESCRIBE THE PAGE]
The main sections it needs, in order: [LIST THEM]
Any must-have elements: [NAV ITEMS, FORM, HERO, FOOTER LINKS, ETC.]

Give me:
1. A complete, valid HTML5 skeleton for this page using semantic
   tags, with placeholder text/labels I can replace.
2. Comments marking each region so a beginner can see the structure.
3. Explicit width/height (or aspect-ratio) on any image/embed slot
   so the layout doesn't shift as things load.
4. A short note on which ARIA attributes (if any) this structure
   genuinely needs — and which it doesn't, because native semantic
   tags already cover them.
5. The one accessibility check to run before I style it.
```

## Example output
[TO FILL AFTER TESTING]

## Tweaks
- Pair with Fluid CSS System next — this gives the bones, that gives the skin
- Ask for the same skeleton "as an Astro/React component" if that's your stack
- Run the output through a validator (validator.w3.org) — semantic doesn't mean valid until checked

---

---
name: Fluid CSS System
category: coding
works_with: [claude, chatgpt, gemini, cursor]
difficulty: plug-and-play
version: 1.0
source: adapted (viral "web architect" prompt pack, rewritten; guru naming and puffery removed)
tested: no
---

# Fluid CSS System
## What it does
Writes a modern, responsive stylesheet from your colors and fonts — custom properties, CSS Grid and Flexbox, `clamp()` for type that scales smoothly, and a working dark mode — applied to semantic tags, no framework required.

## When to use it
- You have an HTML skeleton and need it to look good and scale to any screen
- You want dark mode done properly with variables, not a hacky override
- You're tired of fighting breakpoints and want type/space that flows

## The skill
```
Act as a CSS developer who writes clean, modern, framework-free
styles. Rules: use CSS custom properties for all colors, spacing,
and fonts; use clamp() for fluid font sizes and spacing so fewer
breakpoints are needed; layout with Grid and Flexbox; implement
dark mode with a [data-theme] or prefers-color-scheme approach
using the same variables; mobile-first. Keep specificity low.
Don't invent performance numbers — just write maintainable CSS.

My colors (name + hex): [LIST, or "suggest an accessible set"]
My fonts (heading / body): [NAMES, or "system fonts"]
The vibe: [E.G. calm/technical, warm, high-contrast]
The HTML I'm styling: [PASTE SKELETON, or describe the sections]

Give me:
1. A :root block of custom properties: color, spacing scale, font
   sizes with clamp(), radii — the design tokens.
2. A dark-mode block that reuses those variables (no duplicated
   rules).
3. Base + layout styles applied to the semantic tags, using Grid
   for page structure and Flexbox for components.
4. One fluid type scale so headings and body scale between a min
   and max without a wall of media queries.
5. A contrast note: confirm my text/background pairs meet WCAG AA,
   and flag any that don't with a fix.
```

## Example output
[TO FILL AFTER TESTING]

## Tweaks
- Give it real hexes and it returns your palette; give it "suggest" and it proposes one — either works
- Ask "add a print stylesheet" or "add reduced-motion handling" as a follow-up; both are one line for it
- Verify the AA contrast claim yourself with a checker — never trust a color-contrast claim unverified

---

---
name: Landing Section Layout
category: coding
works_with: [claude, chatgpt, gemini, cursor]
difficulty: plug-and-play
version: 1.0
source: adapted (viral "web architect" prompt pack, rewritten; conversion puffery removed)
tested: no
---

# Landing Section Layout

## What it does
Builds a repeatable, scannable section block for a landing page — clear heading hierarchy, readable at a glance, consistent spacing — so a visitor understands what you offer without hunting. Structure and clarity, not dark-pattern "attention capture."

## When to use it
- You have a page skeleton and need the middle sections (services, features, products) laid out
- Your current sections are a wall of text nobody reads
- You want one reusable pattern instead of designing each section from scratch

## The skill
```
Act as a front-end developer who values clarity over cleverness.
Rules: one clear idea per section; a plain, descriptive heading
(not vague hype); readable line lengths (~60-75 chars); real
contrast for legibility, not just for shouting; consistent vertical
rhythm between sections. No manipulation patterns, no fake urgency,
no "forces the user to act" language — just a section a tired person
can scan and understand.

What this section presents: [SERVICES / FEATURES / PRODUCTS / STEPS]
The items to show: [LIST EACH with a one-line description]
My existing CSS variables (if any): [PASTE, or "none yet"]

Give me:
1. A reusable semantic HTML pattern for one section, filled with my
   items — heading, supporting line, and a clean grid/list of the
   items.
2. The CSS for it, using my variables (or sensible defaults),
   responsive with Grid, no fixed heights that cause overflow.
3. A clear heading hierarchy: how this section's h2/h3 relate to
   the page's h1, so structure stays logical.
4. The "repeat it" note: exactly what to change to reuse this
   pattern for the next section, so the page stays consistent.
5. A readability check: line length, contrast, and tap-target size
   on mobile.
```

## Example output
[TO FILL AFTER TESTING]

## Tweaks
- Keep one idea per section — the instinct to cram three is what made the old page unreadable
- Reuse the exact pattern down the page; consistency is what makes a site feel designed
- All-caps for a short label is fine; all-caps for a sentence hurts readability — keep it to 1-3 words

---

---
name: Lead Capture Form
category: coding
works_with: [claude, chatgpt, gemini, cursor]
difficulty: plug-and-play
version: 1.0
source: adapted (viral "web architect" prompt pack, rewritten; "friction funnel" framing removed)
tested: no
---

# Lead Capture Form

## What it does
Builds an accessible, low-friction contact or signup form as a proper semantic component — labels, validation, and a clear success state — that asks for the least it can and actually works for keyboard and screen-reader users.

## When to use it
- You need a contact, callback, or email-capture form on a page
- Your current form is unlabeled, ugly on mobile, or inaccessible
- You want the markup right before wiring it to a backend

## The skill
```
Act as a front-end developer who builds accessible forms. Rules:
every input has a real <label> (not just a placeholder); use correct
input types and autocomplete attributes; mark required fields
honestly; validation messages are announced to assistive tech;
the form works and looks right on a phone; ask for the minimum
fields the goal needs — no data-harvesting for its own sake.

What the form is for: [CONTACT / CALLBACK / EMAIL SIGNUP / OTHER]
The absolute minimum fields the goal needs: [LIST — challenge me if
I list more than necessary]
Where it submits: [an endpoint / mailto / "just the markup for now"]
My CSS variables (if any): [PASTE or "none"]

Give me:
1. The semantic HTML form: labelled fields, correct types,
   autocomplete, required flags, and an accessible submit button.
2. Inline validation: HTML-native where possible, with clear,
   polite error text and an aria-live region for messages.
3. The success state: what the user sees after submitting, handled
   accessibly (focus moves to the confirmation).
4. CSS using my variables, comfortable tap targets, no layout shift
   when errors appear.
5. A one-line note on where to add spam protection (honeypot /
   provider-side) without adding a captcha wall unless truly needed.
```

## Example output
[TO FILL AFTER TESTING]

## Tweaks
- Fewer fields = more completions and less to store; let it talk you down to name + one field where it can
- Ask for the matching submit-handler in your stack (Netlify Forms, Formspree, an API route) as a follow-up
- Test it with the keyboard only — tab through, submit, read the error. If that works, everyone can use it

---

---
name: Frontend Performance Audit
category: coding
works_with: [claude, chatgpt, gemini, cursor]
difficulty: plug-and-play
version: 1.0
source: adapted (viral "web architect" prompt pack, rewritten; vanity-metric framing removed)
tested: no
---

# Frontend Performance Audit

## What it does
Takes your real performance numbers (from Lighthouse or PageSpeed Insights) and your page, and tells you the specific thing that's actually slow — and the concrete fix — instead of a generic "optimize your images" list.

## When to use it
- Your page feels slow and you have (or can get) real metrics
- Lighthouse flagged something and you don't know what to change
- Before launch, to catch the obvious wins

## The skill
```
Act as a web performance engineer. Rules: work only from the real
numbers and code I give you — do NOT invent metrics or guess at
values I didn't provide; if you need a number I don't have, tell me
how to get it (Lighthouse / PageSpeed Insights / DevTools). Rank
fixes by likely impact, and be honest when a "problem" is actually
fine. No made-up percentages.

My real metrics (paste what you have — Lighthouse/PSI): [LCP, CLS,
INP/TBT, total page weight, request count — whatever I actually have]
What the page is: [DESCRIBE, and paste the <head> and any big
scripts/styles if you can]
What feels slow to me: [FIRST PAINT / JANK ON SCROLL / SLOW TAP /
JUST THE SCORE]

Give me:
1. The diagnosis: which metric is actually bad vs which is fine,
   read from MY numbers — no invented data.
2. The likely cause in my specific page (render-blocking resource,
   unsized image causing CLS, heavy script, no caching, etc.),
   pointing at the actual code I pasted where possible.
3. The fix list, ranked by expected impact, each with the concrete
   change to make — not "optimize images" but "add width/height to
   the hero img and serve it as WebP".
4. What to re-measure after each fix to confirm it worked.
5. The honest note: anything that's already good and not worth
   touching, so I don't waste effort chasing a green score.
```

## Example output
[TO FILL AFTER TESTING]

## Tweaks
- Get real numbers first: run the page through PageSpeed Insights and paste the results — the audit is only as good as the data
- Fix one thing, re-measure, then the next; bulk changes hide which one helped
- A perfect score isn't the goal — a page that feels fast to real users is; the skill will tell you when to stop

---

---
name: Ship a Landing Page
category: coding
tools_required: [Any AI chat (or Cursor), a code editor, a free host like Netlify or Cloudflare Pages]
cost: free
time_to_set_up: an afternoon for a real page
version: 1.0
source: original system built from 5 adapted skills (see /library coding skills)
tested: no
---

# Ship a Landing Page

## What it does
Chains five skills into one clean path from empty folder to a live, fast, accessible landing page: bones, skin, sections, a form, and a performance pass — no framework required, no template-shop markup.

## What you need
- A text editor (or Cursor) and any AI chat
- Your content: what the page is for, the sections it needs, your colors/fonts
- A free static host (Netlify drop, Cloudflare Pages, GitHub Pages)
- The five skills: Semantic HTML Skeleton, Fluid CSS System, Landing Section Layout, Lead Capture Form, Frontend Performance Audit

## The workflow (in order)
1. **Semantic HTML Skeleton** — get the page bones right first: semantic tags, images sized so nothing jumps, accessible structure. Save as index.html.
2. **Fluid CSS System** — feed it your colors and fonts; get design tokens, fluid type, Grid/Flex layout, and a real dark mode. Save as style.css.
3. **Landing Section Layout** — build one reusable, scannable section pattern, then repeat it down the page for each block (services, features, proof). Consistency here is what makes it look designed.
4. **Lead Capture Form** — add the accessible contact/signup form, asking for the minimum. Wire it to your host's form handler or an endpoint.
5. **Deploy, then Frontend Performance Audit** — put it live (drag the folder to Netlify), run it through PageSpeed Insights, paste the real numbers into the audit, fix the top item, re-measure.

## Maintenance habits
- Keep the CSS variables as your single source of truth — restyle by changing tokens, not chasing hex codes through the file.
- Re-run the performance audit after adding anything heavy (a video, a third-party script).
- Every image gets width/height (or aspect-ratio) forever — it's the cheapest way to keep the layout from jumping.

## What we stripped
This was rebuilt from a pack of eight "elite web architect" prompts drowning in invented jargon: "Asymmetric-Yield Semantic Skeleton", "Prestige-Identity Fluid Grid", "Pattern-Interrupt", "Humanized-Velocity Value Engine", "Skeptic-Filter Trust Shield". None of the coined names mean anything — they're the exact "rename a normal thing to sound proprietary" move this library exists to undo. Stripped: the manufactured terminology, the "immediate visual authority", "forces immediate interaction", "dissolving visitor skepticism", "ultra-premium launch" puffery, and "eliminate layout shift in 30 minutes" style promises. Also dropped from the original eight: two prompts that were pure conversion theater (a "proof-of-concept matrix that forces interaction" and a vague "3 interactive script layouts") with no real technique under them, and one that was just "write landing copy" — already covered by the content skills. What survived is five genuine front-end tasks. The techniques under them (semantic HTML, clamp() fluid type, Grid/Flexbox, dark-mode custom properties, Core Web Vitals) are current, verifiable best practice — not secrets, just craft.
