Are you BotReady?
A billion people a week now ask an assistant instead of searching. An assistant can only recommend a site it managed to read. We fetch yours as ChatGPT, Claude and Perplexity do, show you exactly what each one got back, and write the files that fix it. Thirty seconds, free.
- sites checked so far
- 218
- sites checked so far
- average score out of 100
- 62
- average score out of 100
- scores a D or worse
- 1 in 3
- scores a D or worse
- refused our crawler outright
- 33
- refused our crawler outright
Right now an assistant is recommending someone else
ChatGPT passed a billion people a week in August. Every one of those answers is built from pages the assistant managed to fetch, and if it never reached yours your name never comes up — with nothing anywhere to tell you it happened.
- Analytics cannot show you a visit that never happened.
- Usually it is a Cloudflare default nobody switched on deliberately.
- Eight files fix it. Five you upload, three you fill in.
×5
- ···Chrome 128waiting
- ···ClaudeBot/1.0waiting
- ···GPTBot/1.2waiting
- ···PerplexityBotwaiting
- ···Google-Extendedwaiting
Crawler user agents we see in the wild: ClaudeBot/1.0, GPTBot/1.2, PerplexityBot/1.0, Google-Extended, Applebot-Extended, Bytespider, CCBot/2.0, Amazonbot, cohere-ai, Meta-ExternalAgent, Diffbot, YouBot.
Same URL, same second, two different answers
Issues, roadmap and a weekly plan in one place. Free for your first five people.
GET / HTTP/1.1 Host: yoursite.com User-Agent: ClaudeBot/1.0 HTTP/1.1 403 Forbidden server: cloudflare cf-mitigated: challenge (nothing readable)
Same URL, same second, two clients. Chrome got 9,240 characters of your pitch. ClaudeBot got a challenge page and gave up.
Three ways this quietly costs you customers
None of them show up in a dashboard, which is why they run for months.
The shortlist forms without you
Ask any assistant about your category and you get about three names back. We fix the one thing keeping you off that list — whether it can read your site at all.
Nobody chose the block
One Cloudflare default turns away every AI client at once, and it has been running quietly since the day somebody enabled it. Thirty seconds tells you whether that is you.
The fix is 8 files
We write them from your own pages. Five are ready to upload; three carry the questions an assistant asks, with the answers left for you.
Thirty seconds, start to finish
Paste your URL
Paste a URL. We fetch it as Chrome, ClaudeBot, GPTBot, PerplexityBot and Google-Extended.
See what each one got
You get a grade, six category scores, and a plain list of what is blocking you.
Ship the fix pack
Eight files, written from your own pages. Upload, fill in the blanks, re-run.
- AI clients requested per scan
- 5
- AI clients requested per scan
- checks, each with a published weight
- 21
- checks, each with a published weight
- from URL to a shareable result page
- 30s
- from URL to a shareable result page
- code you have to write
- 0
- code you have to write
Not every AI crawler in your logs is the one it says it is
A user-agent is a claim anybody can type. Point your access logs at us and every fetch gets checked against the vendor’s own published addresses and reverse DNS, so a real crawler and a scraper wearing its name stop counting as the same visit.
Your visitors’ addresses are never written down — not stored, not hashed. A crawler’s lives just long enough to ask DNS whether the claim is true.
An illustration of the four verdicts, not a customer’s data.
Find out in thirty seconds
Free, and the result is a page you can send straight to whoever owns the site.
Run the free check