AuditLampAudit is the crawler AuditLamp uses to read a website and audit it. It runs only when that site's owner asks us to, by submitting the URL. An audit reads a small, bounded set of public pages plus robots.txt, sitemap.xml and llms.txt, and scores what it finds against 164 graded checks: SEO, AEO and GEO plus the Vibe Audit (security, accessibility, privacy).
How it identifies itself
Every request AuditLamp makes carries a user-agent that names us and links back to this page, so you can always tell it apart in your logs:
The browser portion is kept so we can read sites that block unfamiliar clients outright; the AuditLampAudit/1.0 token is the part that identifies us. Operators can switch AuditLamp to a fully transparent, non-browser user-agent if they prefer.
How to allow or block it
AuditLampAudit obeys robots.txt. To control it by name, target the AuditLampAudit token. For example, to block it entirely:
We honour Disallow rules for any page we discover while crawling, and we honour a Crawl-delay you set. The one page we always read is the exact URL the owner submits to us for an audit, because that is a direct, owner-initiated request for that page.
How politely it crawls
- Requests to your site are spaced out and capped in number. A single audit reads a small, bounded set of pages, not your whole site at speed.
- We back off automatically when a server returns a rate-limit or temporary-error response.
- We fetch public pages only, the way any browser would; we never log in, submit forms, or follow links to other people's sites.
Questions or problems
If AuditLampAudit is causing trouble, or you want it to stop, email hello@auditlamp.com and we will sort it out quickly.
Policy help: AI bot policy templates · how the scan works.