Back
Site Audit
Crawl a site, audit it, and let AI read the findings back
A marketing workbench that crawls a site and reports what is wrong with it — robots and sitemap handling, page structure, metadata — then puts an AI over the findings so the write-up comes back in words a client can read. Work is organised into projects, every run is kept in history, and usage is metered per account.
Web app Shipped product · running
Key Features
Site audit over robots.txt, the sitemap and the page structure
AI turns the raw findings into a report a client can read
Work grouped into projects, with every run kept in history
Charted results rather than a wall of rows
Per-account usage metering, with an admin view over all of it
Email delivery of the finished report
Built with
Next.jsTypeScriptSupabaseAnthropicCheerioRecharts
How it works
Next.js route handlers do the crawling with Cheerio and a proper robots parser, bounded by a concurrency limit so an audit cannot hammer the site it is reading. Supabase holds the projects, analyses and usage counters; the AI pass runs over the stored findings rather than over the live page, so a report can be regenerated without crawling again.