Website analysis

Extract a site's visual DNA.

Turn a public website into a detailed reference ZIP, then use it with Claude, Codex, OpenCode, Cursor, and any other AI agent to build a new site in the same visual style.

Up to 10 pages per run Hosted captures when available

INSIDE A REFERENCE PACK
website-analysisZIP
README.mdGUIDE
data/pages.jsonJSON
design-tokens.mdNOTES
screenshots/WHEN CAPTURED
STRUCTURED FOR THE NEXT BUILD
01 Discover key pages 02 Record design details 03 Download a structured pack

AI coding workflows

Give an AI agent the site's visual DNA.

The comprehensive ZIP brings together page structure, content, design tokens, typography, layout, responsive behavior, imagery, and captures when available. Start a brand-new site with Claude, Codex, OpenCode, Cursor, and any other AI agent that can read project files. The pack gives the agent a clear reference for the source site's style, art direction, and overall feel.

ClaudeCodexOpenCodeCursorAnd any other AI agent

What the pack contains

Page details, design records, and available captures.

Bring structure, visual observations, and content notes into one organized set of files for designers, developers, and AI-assisted rebuild workflows.

Page structure

See the page set alongside its navigation, heading structure, section order, and selection notes. Each observation stays linked to the page it describes.

Design observations

Carry colors, typography, layout patterns, responsive changes, media references, and visible control cues into the next design or implementation conversation.

A portable reference pack

Get Markdown guides, JSON records, theme files, and images from successful captures in a ZIP assembled in your browser.

Inside the reference pack

One archive captures a site's visual DNA.

Move from an overall map to page evidence, design tokens, responsive details, and source captures. Give the same organized reference to a designer, developer, Claude, Codex, OpenCode, Cursor, or any other AI agent for a new build.

Page map

See discovered pages, navigation relationships, and the recorded selection decisions for the run.

Page notes

Keep visible headings, content, and section order in context with the page where they were observed.

Design details

Review tokens, typography, layout samples, media references, and responsive behavior in written and structured form.

Available captures

Use desktop, mobile, and section images as visual evidence when those captures complete successfully.

Analysis workflow

Four steps from a URL to a ZIP.

STEP 01

Start with a public URL

Choose a public site. Site Stripper looks across relevant same-site links and ranks useful pages for a bounded run.

STEP 02

Review the observations

Explore the page map, visible content, design records, responsive notes, coverage details, and available captures together.

STEP 03

Build from the visual DNA

Give the ZIP to Claude, Codex, OpenCode, Cursor, or any other AI agent. Its detailed notes and design tokens provide a strong starting point for a fresh site in the reference's style.

Share source notes with the build team.

Website rebuilds often begin with scattered screenshots, copied links, and notes. Site Stripper brings the source site's visual DNA into a comprehensive package that a design and development team or an AI coding agent can use as a shared starting point.

Keep the source in viewPage notes and structured records identify where the observations came from.
Compare more than appearanceContent, navigation, layout, responsive patterns, and available images sit alongside one another.
Start a new buildUse the package with Claude, Codex, OpenCode, Cursor, and any other AI agent to carry the same visual language into a brand-new website or web app.
A sample interior architecture website homepage with project photography and visible page structure
Example source site, desktop view

Start with the source

Keep the page, its content, and the design details together.

A strong reference begins with what is actually on the page. Site Stripper organizes visible content, page relationships, design observations, and available captures so they can be reviewed in context.

  • Follow the page map into specific source pages.
  • Compare visual details with structured records.
  • Use the ZIP as reference material for a separate build.
See the sample pack

Sample output

Inspect the files in a reference pack.

Explore representative documentation, data files, and capture paths from the package generator.

View the sample pack

Limits and coverage

Know what each run includes.

How many pages can one run analyze?

Hosted runs can analyze up to 10 selected pages in up to 150 seconds.

Are screenshots included in every pack?

Capture files are included when available. Source-site behavior and hosted limits can make captures partial or unavailable.

Does Site Stripper rebuild the website?

No. It gathers reference material to guide a separate design and development workflow.

Get started

Create a Site Stripper account.

Use the account form to begin with page structure, design observations, and organized output files.

Create your account