Paste a URL. See how machines see you.
AuditLamp grades how easy it is for search engines and AI systems to find, trust, and use your public website, then ranks the technical fixes that matter most. You do not need to learn AEO or GEO first. You need a score, five plain pillars, and a fix list.
Whole site up to 150 pages · 164 graded checks · no email on the diagnosis
Three steps. Then you act.
Fetch
We request your public pages the way careful crawlers do: real robots.txt, real HTML, named AI bot probes, sitemap and llms.txt when present. Public URLs only. No login required to see the free diagnosis.
Grade
We run 164 graded checks: classic search signals, answer and AI visibility, local and product markup where they apply, plus the Vibe Audit for security, accessibility, and privacy. Each check is named and sourced, not a mystery score.
Rank
You get an overall grade, worst problems first, and plain-English fixes. Free is the full diagnosis. The $10 pack is the file you keep (PDF, action plan, AI fix file). Monitor watches what changes after deploys.
Learn the pillars. Open the checks when you need them.
Findability
Can crawlers reach and index you? Robots, sitemap, index signals, soft 404s, host consistency.
AI visibility
Can major AI crawlers read you when you want them to? Bot matrix, live reachability, CDN vs robots conflicts, answer-friendly structure, optional llms.txt.
Technical SEO
Titles, headings, canonicals, structured data, content depth, entity and schema quality.
Trust and experience
HTTPS and security basics, privacy signals, accessibility, page weight, field speed when available.
Local and commerce
When your site sells a place or a product: NAP consistency, local schema, Product and offer markup, merchant fields.
Full catalog: what we check. Primary-data studies: AI crawler blocking study.
AI tools do not magically know your brand.
Like Google, AI products send crawlers to read public pages. Your robots.txt file, and sometimes your CDN or firewall, acts like a guest list: each bot can be allowed or blocked by name. Training bots and search or answer bots are often different. Blocking training is a policy choice. Blocking the bots that power answers can make you invisible when people ask ChatGPT or Perplexity who to hire or what to buy, even if Google SEO looks fine.
We report which important AI-related crawlers are allowed or blocked, and whether a live probe with that bot identity reaches the page. We do not decide your policy. We do not promise that allowing a bot will make a model cite you. We prove whether anything on your side is closed.
Deeper: GPTBot blocked, what we check · robots say yes, CDN may still block.
Readiness, not a citation guarantee.
Not a ChatGPT rank tracker. Tools like Peec or Profound monitor whether models mention your brand. We diagnose whether your site can be read, trusted, and quoted in the first place.
Not a promise of rankings or AI citations. Models change. Personalization exists. We measure eligibility and machine-readiness on your public pages.
Not a penetration test or full content strategy. Snapshot of public surface, up to 150 pages per free scan, with documented caps on pricing.
Not folklore. If we cannot source a check, we do not ship it. That is the honesty brand.
Free shows the wound. Paid is the file and the watch.
Free Visibility Scan: whole-site diagnosis up to 150 pages, ranked fails, plain English. No email required to see it.
Full Report $10 once: the pack you keep. PDF, prioritized action plan, AI-ready fix file. Not "unlock hidden results." Packaging and handoff.
Monitor $29/mo: re-checks and alerts when access or scores move after deploys. Continuity, not more mystery checks.
Agency $99/mo: five sites, client work, same engine. Details on pricing.
Short answers.
What does my grade mean? Overall site health from graded checks. Start with Critical and gate fails. A strong Google ranking can still hide AI bot blocks.
Why so many checks? 164 is credibility for experts. Your default job is the top ranked fails, not reading every row.
Will this hurt my site? Bounded crawl of public pages when you request a scan. User-agent is documented on our crawler page.
Do I need llms.txt? Helpful experiment, not a magic ranking file. Missing it should not be your only panic. Access and extractable content matter more.
Who is this for? Site owners, designers, publishers, local and ecommerce, AI startups, SaaS, agencies verifying client work. Vertical doors: local, ecommerce, AI startups, designers.