Agent Analytics

See which AI bots read your siteand which pages they cannot read

An AI answer starts with a crawl. FullReach AI shows which AI crawlers visit your site, which requests fail, what your robots.txt blocks and how many visitors AI sends you.

Free. No account. See which AI crawlers your site lets in. Or start a free trial

Pipedrive · Crawler VisitsServer logs · Last 28 days
Crawler requests30,218
from 5 AI crawlers
Verified97.6%
against the operators’ published lists
Failed1.8%
403, 429 and 5xx answers
Left out679
656 spoofed, 23 probes
Requests per dayGPTBotOAI-SearchBotChatGPT-UserPerplexityBotClaudeBot
CrawlerPurposeRequestsFailed
GPTBotOpenAITraining8,4201.2%
OAI-SearchBotOpenAISearch7,7600.8%
ChatGPT-UserOpenAIAsked by a person5,9282.6%
PerplexityBotPerplexitySearch4,4304.9%
ClaudeBotAnthropicTraining3,6800.4%

Illustrative sample. The brands are real. The figures in the sample panels are invented.

Crawler visits

Each request from a known AI crawler, by day, crawler and page.

Verification

Each request’s address against the list its operator publishes: verified, spoofed or unverifiable.

Crawlability

What your robots.txt allows for about 30 AI crawlers, and what a plain request to your homepage gets.

AI visits

The sessions that AI assistants send to your site, read from your Google Analytics 4.

Features

Find out if AI can read you

AI cannot cite a page that it cannot read. These checks show you where the crawl breaks.

Connect your logs your way

Upload a log file, send logs with an ingest key, or connect Cloudflare. On an upload, only the crawler rows leave your browser.

Example
Upload a log fileCombined Log Format, CSV or NDJSONSet up →
Send logs with an ingest keyA Vercel drain, Akamai DataStream or JSONConnected
CloudflareA read-only token. 30 days back, then dailySet up →
Google Analytics 4Our service account as a ViewerConnected

Tell real crawlers from fakes

Anyone can call a bot GPTBot. FullReach AI checks each request’s address against the list that the operator publishes. Spoofed requests stay out of every count.

Example
GPTBot98.9%212 spoofed
OAI-SearchBot99.4%0 spoofed
ChatGPT-User97.1%38 spoofed
PerplexityBot91.3%406 spoofed
VerifiedUnverifiable

Catch the pages that fail

See every page that answers an AI crawler with 403, 429 or 5xx. Pages not found are listed apart.

Example
403/pricingPerplexityBot
429/blog/crm-guideChatGPT-User
404/features/old-pageGPTBot

Audit your robots.txt

See what your robots.txt allows for each AI crawler, and the line that decides it. Test a URL, and compare the last 30 versions of the file.

Example
GPTBotAllowed
OAI-SearchBotAllowed
PerplexityBotAllowed
Google-ExtendedBlocked, with exceptions
CCBotBlocked

Check your homepage every day

FullReach AI checks that your homepage reads without JavaScript, and asks ChatGPT’s own fetcher to open it and quote it.

Example
A plain request opens it
It shows its text without JavaScript
It carries no AI opt-out directive
We could not confirm that ChatGPT can read it

Join crawls to citations and visits

Own Pages puts each page’s crawler requests next to the answers that cite it and the visitors AI sends to it. A page that AI crawlers read but no answer cites stands out.

Example
PageCrawler requestsCited inAI sessions
/2,24941 answers812
/pricing1,64728 answers455
/blog/best-crm-small-business1,891Never cited37
/features/pipeline-management1,5529 answers120

AI sessions come from Google Analytics 4. The count is a floor, because GA4 files some AI visits under Organic Search or Direct.

MCP

Ask ChatGPT which AI bots read your site.

Connect FullReach AI to the assistant your team already uses. It reads the same figures as your dashboard, and the answers behind them.

ChatGPTClaudeClaude CodeCursor
See the MCP server
Example
ChatGPTFullReach AI connected

Is anything stopping AI crawlers from reading our site?

get_crawler_visitsget_crawlability

Your robots.txt blocks CCBot. It blocks Google-Extended, with exceptions. PerplexityBot got 217 failed requests in the last 28 days, most of them a 403 on /pricing.

Today’s check could not confirm that ChatGPT can read your homepage.

FAQ

Agent Analytics questions

Which AI crawlers do you recognise?

About 30 AI crawlers that their operators document. Among them are GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot and ClaudeBot.

Do I need to install a script?

No. Upload a log file, send logs from a CDN or a server with an ingest key, or connect Cloudflare with a read-only token.

What does “spoofed” mean?

The request used the name of a known AI crawler, but it came from an address outside the list its operator publishes. Cloudflare sends no addresses, so its requests stay unverifiable.

Do you store my visitors’ IP addresses?

No. On an upload, only the crawler rows leave your browser. FullReach AI reads the address to verify a request, then keeps daily counts without it.

Why are the AI visit counts a floor?

GA4 cannot attribute every AI visit. It files some under Organic Search or Direct. FullReach AI shows what GA4 counts and adds nothing to it.

Can I check my site without an account?

Yes. The free AI crawler checker reads the robots.txt and the homepage of any website. It shows which AI crawlers may read it.

Free trial

Learn what AI says about you.

Your first analysis runs right away.

7-day free trial · No credit card required

See pricing