llms.txt for AI assistants
A file at /llms.txt: a short Markdown map of the site written for answer engines and AI assistants, so they describe it from what you wrote rather than from what they inferred. A convention, not a standard — nothing requires it.
/llms.txt is a proposed convention for helping large language models understand a site. It is a Markdown file in the site root: an H1 with the site name, a one-paragraph summary in a blockquote, then sections of links to the pages that matter, each with a short description. AI assistants and answer engines that read it can describe your site from your own words rather than piecing it together from navigation menus and cookie banners.
It is not a standard, and no search engine has promised to use it — nothing requires it, and the report never treats its absence as a problem. It is cheap to add, though, and a sensible part of AI visibility.
How SiteScan checks
The scan requests /llms.txt without following redirects and counts it as published when the answer is a 200 with a non-empty body that is not HTML or XML — so a single-page app that returns its index for every path is not mistaken for one. When found, the row shows the file size in bytes.
How to write one
Keep it short and current: what the site or product is, who it is for, and links to the key pages (documentation, pricing, contact) with one line each. Serve it as text/plain or text/markdown. For a complete copy of your content in one file, see llms-full.txt.
See this check on your own site
Run a free scan: the report shows this row for your domain along with the other checks of the site audit.
More in “Well-known files”
What SiteScan checks
TLS certificate
- TLS certificate expiry
- Outdated TLS versions (TLS 1.0/1.1)
- TLS certificate chain
- TLS certificate issuer
- Certificate signature algorithm
Domain registration
- Domain registration expiry
- Registered domain name
- Domain registrar
- Domain registration date
- Domain last updated date
- Domain status codes (EPP)
- Domain transfer lock
- Registrar abuse contact
- RDAP and WHOIS data source
Indexing
- robots.txt file
- CSS and JS blocked in robots.txt
- XML sitemap
- Sitemap index file
- Off-domain URLs in the sitemap
- Sitemap lastmod dates
- Broken URLs in the sitemap
- RSS and Atom feed
Addresses & redirects
Security headers
- HTTP security headers and grade
- Strict-Transport-Security (HSTS) header
- Content-Security-Policy (CSP) header
- X-Content-Type-Options: nosniff
- X-Frame-Options and clickjacking
- Referrer-Policy header
- Permissions-Policy header
- HSTS preload list
- CORS and cross-origin reads
- HTTP TRACE method
Delivery & reputation
- Cache-Control header for HTML
- ETag and Last-Modified revalidation
- HTTP/3 and QUIC support
- Server version disclosure in headers
- DNS blocklist (DNSBL) check
Internal links
Exposure
- Exposed sensitive files (.env, .git, backups)
- Directory listing enabled
- Public JavaScript source maps
- Exposed WordPress endpoints
Well-known files
- security.txt file
- llms.txt for AI assistants
- llms-full.txt file
- ads.txt authorized sellers file
- humans.txt file
- Well-known change-password URL
DNS & email
- MX records (mail servers)
- SPF record
- DMARC record and policy
- DNSSEC signing
- CAA record
- Nameserver redundancy
- IPv6 support (AAAA record)
- MTA-STS policy
- TLS-RPT reporting record
- BIMI record (brand logo in email)
- Wildcard DNS record
- Dangling CNAME and subdomain takeover
Privacy & consent
- Analytics and tracking scripts
- Cookie consent banner
- Privacy policy link
- Terms of service link
- Imprint (Impressum)
- Accessibility statement
Core Web Vitals
- Lighthouse performance score
- Largest Contentful Paint (LCP)
- Cumulative Layout Shift (CLS)
- Interaction to Next Paint (INP)
- CrUX field data vs lab data