What the tech stack check tells you
Websites leave traces of what they are built with: a generator tag from the CMS, script paths from a framework, cookies set by a shop platform, headers added by a CDN, and DNS records pointing at a hosting or email provider. The tech stack check collects those traces from the site's home page and its DNS, matches them against our own catalog of fingerprints, and lists every technology it finds with the evidence behind it.
What we check
- Headers. Every response header on the way to the home page, including the redirects. Server names, CDN request IDs and platform headers are some of the clearest signals. We read at most 1 KB of each value.
- Cookies. The names of cookies the site sets, matched against patterns such as
wordpress_logged_in_*or_shopify_y. We never keep or show the values. - HTML. Generator meta tags, script and stylesheet addresses, attributes on the html element and the ids frameworks give their root elements, plus a search for fixed markers in the first 512 KB of the page.
- DNS. Up to eight lookups: the domain's name servers, mail servers and TXT records, the CNAME of the site and its www name, its address and the reverse name of that address. Name servers tell us the DNS host, CNAMEs and reverse names the web host, mail servers the email provider, and verification tokens in TXT records the services a domain has signed up for.
Versions are read only where the site publishes them, for example in a generator tag or a server header. When a site shows software versions we add an information finding, since visible versions make it easier for attackers to look up known flaws.
How confidence works
Each fingerprint carries a weight for how much it proves. When several signals match, their weights combine the way independent clues do: two strong hints together are close to certain. A technology that implies another adds it too, so WordPress brings PHP and Next.js brings React, at the same confidence. We show technologies at 50% or more, and mark anything below 80% as likely.
Why the evidence matters
Fingerprints can be wrong. A theme can keep an old generator tag after a site moves platform, a CDN can sit in front of a different host, and a script can be copied from one site to another. That's why every entry shows the signals behind it rather than a bare name. If an entry surprises you, the evidence tells you where to look: a header to remove, an old file still linked from the page, or a DNS record left over from a service you stopped using long ago.
Reading the result
Technologies are grouped by category: CMS, online shop, framework, hosting, CDN, DNS host, email provider, web server, language, JavaScript libraries, analytics, tag managers, consent platforms, chat, payments, security services, fonts and services found through DNS verification. Each entry shows its version when known, its confidence in words and a collapsed list of the evidence, such as "meta generator" or "NS records". Vendor names link to their own sites.
The check has no score: what a site runs isn't good or bad in itself. When the home page blocks our checker we still answer from DNS, mark the result partial and say so. When a DNS lookup doesn't answer, the result is also partial and notes it.
What the check doesn't do
We fetch only the home page, with an honest user agent, and we don't run JavaScript, so tools loaded later by a tag manager may not appear. We don't scan other pages, ports or files. The network owner of the site's address, which some tools show, waits for our own routing data rather than a third-party lookup.