Create LLM Prompt specs (md files) and screenshots used for website replication

Create LLM Prompt Package to Replicate Website UI

Overview
I need someone to document an existing website so that an LLM (cursor) can use the output to replicate its UI onto another site. The deliverable is NOT code — it is a structured set of spec files, screenshots, and prompt instructions that I will feed directly into an AI coding workflow.

This is ideal for someone with a strong eye for UI detail, experience writing technical documentation, and familiarity with how LLMs interpret design specs.

### What You Will Deliver

**1. design-tokens.md**
- Every color (hex values), extracted from the site
- Typography: font families, sizes, weights, line heights for each text style
- Spacing scale, border radius values, box shadows
- Breakpoints for responsive layout

**2. component-inventory.md**
For every component on the site (navbar, hero, cards, forms, footer, etc.):
- Layout description (structure, alignment, grid/flex behavior)
- All visible text styles and content patterns
- Responsive behavior at mobile, tablet, desktop
- Any animations or transitions noted

**3. page-map.md**
- List of all pages/routes
- Section order per page
- Shared layout elements (header/footer/sidebars)

**4. interaction-spec.md**
- All interactive states: hover, active, focus, disabled
- Toggle behaviors (e.g. tabs, accordions, dropdowns)
- Form validation states (empty, error, success)
- Scroll-triggered behaviors or animations

**5. replication-prompt.md**
- A master prompt template an LLM can follow to rebuild each component
- Includes instructions on how to reference the other MD files
- Notes any ambiguities or areas needing human judgment

**6. Screenshots Folder**
Organized screenshots for every component and state:
- Desktop and mobile views
- All interactive states (hover, open/closed, error/success)
- Annotated where helpful (arrows, labels, callouts)
- Named to match component names in component-inventory.md