Windows app · in development
See your whole website the way search engines and AI see it.
ScrapeDesk crawls a website from your own computer and turns it into one searchable report: technical SEO, readiness for AI answer engines, the site's design tokens and the accessibility of its colours. Nothing is sent to a cloud service.
Not for sale yet. We are testing it on many kinds of real files and measuring what it does on an ordinary PC. The version, the price and the measured results will be published here when it has passed.
Who it is for
Freelance SEOs, web agencies and front-end developers who audit other people's sites and would rather not hand a client's site map to yet another online tool.
What it does
ScrapeDesk in short
Crawl explorer
Search across every title, URL and page text at once. Finds near-duplicate pages, groups pages that share one template, and lists the stylesheets, scripts, forms, images and PDFs it found.
Links and orphans
A link graph of the site, pages ranked by click depth and incoming internal links, and orphan detection that compares the XML sitemap with the pages the crawl really reached.
Design tokens
Reads the colour palette, type scale and button styles a site uses and exports them as W3C Design Tokens, ready for a design system.
Search and AI readiness
Checks titles, descriptions, canonical tags and robots rules. Tests thirteen AI crawlers against the site's robots.txt, checks the llms.txt file and can draft one from the site's best pages. Lists structured data and every PDF found.
WCAG colour checks
Calculates contrast ratios under WCAG 2.1 (text and non-text) and separates text colours from buttons, borders and purely decorative washes.
One file you keep
Every crawl ends in a single self-contained HTML report you can open, search and send to a client. No account, no server.
You enter a starting address, choose how many links deep to go (0 to 100) and how many pages at most (up to 500), and press start. A page that only appears after JavaScript runs can be rendered with the Microsoft Edge already on your PC. The report is built when the crawl ends.
Your data
It stays on your computer.
The only network traffic is the crawl itself, to the website you typed in. No telemetry, no account, no cloud AI and no API keys.
Honest limits
What ScrapeDesk does not do
- It reads pages a visitor can open without logging in.
- A run covers up to 500 pages; very large sites need several runs.
- It reports what it finds. It does not change your website.
- PDFs it finds are listed so they can be checked in TagSmith-PDF; ScrapeDesk itself does not repair them.
Conversation
Questions, ideas, experience with similar tools
Anyone can read. To join in, sign in with Google or a free account. Your first comment is checked by a person; later ones appear at once.