Analyze https://example.com and teach me how the page was built. Start with layout overview, then HTML structure, CSS layout, and any JavaScript behavior. Depth: curious. Don't copy their text — use placeholders if we rebuild.
| User says | Scope |
|---|---|
| "just explain" | Phases 1–6, no code in repo |
| "help me rebuild it" | Full workflow + simplified clone |
| "only the hero" | Phase 2–4 on one section |
| "what framework" | Phase 1 gather + Phase 6, short answer |
- Fetch or open URL; record final URL after redirects
- Document title and meta description (if present)
- Landmarks —
header,nav,main,footer,section,article - Repeated patterns — cards, list items, grid cells
- CSS sources —
<link rel="stylesheet">,<style>, inline styles on key nodes - Layout probes — on hero or main container:
display,flex-direction,grid-template,gap,max-width - Breakpoints — resize or read
@mediarules in stylesheets - Fonts —
font-familyonbody,h1 - Scripts — count,
type=module, known bundles (/_next/,react,vue) - Interactions — menu, tabs, modal, form, scroll effects (click/hover if browser available)
| Environment | Recommended approach |
|---|---|
| Cursor + browse/MCP | Navigate URL, snapshot DOM, screenshot regions |
| WebFetch only | Parse returned HTML; note "dynamic content may be missing" |
| User provides screenshot | Describe layout visually; ask for DevTools HTML export for structure |
| Localhost / own project | Full DevTools; tie analysis to their repo files |
| Signal | Likely stack |
|---|---|
#__next, /_next/static/ |
Next.js |
#root + data-reactroot or .react- |
React SPA |
ng- attributes, _ngcontent |
Angular |
v- bindings in HTML |
Vue |
WordPress wp-content, wp-includes |
WordPress theme |
Utility classes flex, grid, text-xl, md: |
Tailwind (or similar) |
| No JS, semantic HTML + one CSS file | Static / hand-written |
- Teach patterns, not stolen assets (logos, marketing copy, licensed fonts)
- Respect
robots.txtand rate limits when fetching - Do not bypass paywalls, CAPTCHAs, or authentication
- For banking, adult, or clearly sensitive sites — confirm user intent
When creating a teaching clone:
| Original | Clone |
|---|---|
| Brand logo | Placeholder box or initials |
| Stock photos | placeholder color block or unsplash generic if user asks |
| 12-section marketing page | 1 hero + 1 card row + footer |
| Complex animation | Static layout first; optional motion later |
## Site overview
- URL:
- Purpose:
- Regions: [ASCII or bullet list]
## HTML structure
- Landmarks:
- Repeating patterns:
## CSS & layout
- Page width / centering:
- Main layout (flex/grid):
- Mobile changes:
## Typography & color
- Fonts:
- Accent:
## JavaScript
- Behaviors observed:
- Framework (if any):
## How you'd rebuild it
1. ...
2. ...
## Terms learned
| Term | One-line |
|------|----------|After analysis → HTML/CSS rebuild:
Using what we learned, build a simplified version of the hero section in my project. Teach each change with DevLearn blocks.
After analysis → curriculum:
/devlearn-curriculum-router— I analyzed a site; what skills should I learn next to build pages like this?