> For the complete documentation index, see [llms.txt](https://aicre8.gitbook.io/aicre8.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://aicre8.gitbook.io/aicre8.com/build-with-aicre8/web-scraping-and-site-cloning.md).

# Web Scraping & Site Cloning

AICre8 can scrape any website and use it as a reference to build your project. Just include a URL in your prompt and AICre8 handles the rest.

### How It Works

When you say something like "Build a site like example.com", AICre8 automatically:

1. Scrapes the reference URL
2. Extracts colors, fonts, layout structure, and content
3. Feeds the scraped data into the AI prompt as context
4. Generates a new site that matches the design language

No extra setup or configuration needed. Works in the IDE chat, via the REST API, and through the MCP server.

### Examples

Try prompts like:

* "Build a site like stripe.com with a dark theme"
* "Clone the layout of openquote.net but use our brand colors"
* "Create a landing page inspired by linear.app"
* "Build a portfolio site like the design of notion.so"

### What Gets Extracted

AICre8 scrapes and analyzes:

* Brand colors and color palette
* Typography and font families
* Page structure and layout patterns
* Hero sections, CTAs, and navigation
* Content and copy (used as reference, not copied verbatim)
* Image descriptions and alt text

### Via the API

Include a URL in your prompt when calling the generate endpoint:

POST /api/v1/projects/:id/generate

Body: { "prompt": "Build a site like <https://example.com> with a dark theme" }

The scraping happens server-side. Costs 1 credit per generation, same as any other prompt.

### Via MCP Server

Use the generate\_code tool with a URL in the prompt:

"Use AICre8 to build a site like stripe.com for my fintech startup"

Your AI agent (Claude Desktop, Cursor, etc.) will call generate\_code and AICre8 handles the scraping automatically.

### Web Search

AICre8 also supports web search to find reference material, competitor sites, and design inspiration. The AI can search the web during generation to inform its output.