Description
Before an AI assistant can recommend your business, it has to be able to read your website. If ChatGPT, Claude, Perplexity or Google’s AI cannot reach your pages, they have nothing to read, and nothing else you do for AI search can help.
Some sites block these assistants without the owner knowing. Security plugins, host defaults and robots.txt rules can all turn AI crawlers away while ordinary visitors see nothing wrong.
I measured 293 local business websites. Eighteen refused an AI crawler while serving browsers normally, and twelve of those eighteen had no robots.txt rule about AI at all, which suggests nobody chose to block them. The full method and numbers are at https://getfoundash.com/research/
This plugin checks, from your own dashboard:
- Whether the agents that fetch your page when a customer asks are being served or refused. OAI-SearchBot and ChatGPT-User are requested live, so you get the HTTP code each one returned
- Whether Perplexity-User and Claude-SearchBot are allowed in your robots.txt. These two are not requested live, so the plugin reports their robots.txt rule and no HTTP code
- Separately, whether the crawlers that collect pages to train on (GPTBot, ClaudeBot, PerplexityBot) are being served or refused, and whether your robots.txt allows Google-Extended, which is a robots.txt setting and not a crawler that visits
- Whether your robots.txt tells any of them to stay out
- Whether WordPress’s own “Discourage search engines” setting is on
- Whether your front page carries a noindex tag telling engines not to list it
- Whether your front page has structured data for an assistant to lift facts from
- Whether you have an llms.txt file
It then gives you one plain sentence saying where you stand.
This plugin checks whether AI can read your site. It cannot tell you whether ChatGPT names your business when a customer asks. Foundash checks that by hand, and the first check is free: https://getfoundash.com/
Questions and answers
AI assistants look for pages that answer a customer’s question plainly. The Questions & answers tab helps you put those answers on your own site:
- Tell it what you do and which town you’re in, and it suggests questions customers often ask, such as “How much does plumbing cost in Reston?”
- Write a short answer to each. It flags answers that are missing, too long, or slow to get to the point.
- Approve the ones you’re happy with. Only approved answers appear on your site, wherever you add the [foundash_qa] shortcode, together with FAQ code built from exactly the same text.
You write the answers. There is no AI service and no account, and nothing is sent anywhere.
How it decides what counts as a block
A refused crawler is not always a deliberate block, so this plugin does not shout about one. It also requests your site as Googlebot as a control. If Google is being refused too, you have a broken bot rule and not an AI policy, and the plugin says so, because that problem is bigger and costs you traffic today.
It only counts 401, 403 and 429 as a deliberate refusal. A 400 is a malformed request and a 500 is your server erroring, and calling either a block would be misleading.
The robots.txt reader follows the actual matching rules. A commented-out line is not a block. A rule written for your site as a whole does not apply to a crawler that has its own rules further down. An Allow: / in the same group as a Disallow: / cancels it. Each of those, done carelessly, invents a block that is not there.
Every result shows the raw HTTP code and the exact robots.txt line behind it, so you can check any result yourself.
Two different crawlers, two different consequences
A single “AI crawlers: blocked” result hides a difference that matters. There are two kinds of AI crawler, and only one of them decides whether you can be recommended.
Some agents fetch your page at the moment a person asks an assistant about your business. If those are refused, the assistant has nothing to read and cannot describe or recommend you. That is worth fixing.
Others collect pages in bulk to train models on. Refusing those is a deliberate choice a lot of sites now make, and on its own it does not stop anyone recommending you.
Across 293 local business websites checked in August 2026, 15 refused an AI crawler in robots.txt, and 14 of those 15 refused only the training crawlers while allowing the search-time agents by name. Reporting those 14 as “blocked from AI” would tell a business it is invisible when it is not.
So this plugin tells you which kind you are refusing, and only raises an alarm for the kind that costs you something.
Why there is no score
This plugin gives you no score out of 100 and no letter grade, on purpose.
Every question it asks has a real answer. A crawler was served or it was refused. A robots.txt rule exists or it does not. Averaging answers like those into a grade throws away the part you can act on, and it makes a short list of plain facts look like a measurement.
You get the facts instead: which crawlers were served, which were refused, the HTTP code each one returned, and the exact robots.txt line behind each result.
What it cannot tell you
Requests come from your own server. If your site sits behind a CDN or firewall and your server bypasses it when calling its own address, a block at that edge will not appear here. Checking from outside your hosting is the only way to be sure.
A slow site gets a partial answer. The whole run is held to a time limit so it cannot outlast PHP’s execution limit and leave you with a blank admin page. If your site is slow to answer requests from itself, the run stops early, the rows it never reached say “Could not test”, and it tells you it stopped.
Google-Extended is reported from robots.txt only. There is no such crawler to refuse. Google reads your pages as Googlebot and then checks the Google-Extended rule to decide whether that content may feed Gemini and AI Overviews. A tool that reports an HTTP status for Google-Extended is reporting a request that never happened.
Privacy
The plugin only requests your own site. It sends nothing to any third party and needs no account, API key or signup. It stores a one hour cache of your last check, plus the questions, answers and business details you enter on the Questions & answers tab. Deleting the plugin removes all of it.
Installation
- Upload the plugin folder to
/wp-content/plugins/, or install it through the Plugins screen. - Activate it.
- Go to Tools > AI Access.
- To add questions and answers, open the Questions & answers tab. Then add a Shortcode block with [foundash_qa] to the page where you want them.
FAQ
Does this send my data anywhere?
No. It makes requests to your own website and nothing else. There is no account, no API key and no tracking.
It says AI crawlers are blocked. How do I fix it?
First check which kind. If the plugin says you are refusing only the training crawlers, nothing is broken and there may be nothing to fix. That is a choice, and it does not stop an assistant recommending you. Only a refused search-time agent costs you anything.
If a search-time agent is being refused, the fix depends on where the block is, and the plugin tells you which.
If it is WordPress’s own setting, go to Settings > Reading and untick “Discourage search engines from indexing this site”.
Worth knowing: since WordPress 5.3 that switch no longer writes Disallow: / into robots.txt. It adds a noindex tag to your pages instead, so crawlers read them and are then told not to list them. A tool that only reads robots.txt cannot see it at all. This plugin checks the setting, the tag and robots.txt separately, and tells you which one is responsible.
If robots.txt is the problem, edit the file or the plugin generating it. If your server is returning 403, the cause is usually a security plugin, a firewall rule, or your host’s bot protection. Cloudflare in particular has a setting that blocks AI crawlers, and it is on by default for some accounts.
Will this get my business recommended by ChatGPT?
No, and be wary of anything claiming it will. Whether an assistant names you depends on reviews, directories, your reputation across the web and how each engine weighs its sources. None of that is under any plugin’s control.
What this tells you is whether you are in the running at all, because a site that cannot be read cannot be recommended.
Does the Questions & answers tab write the answers for me?
No. You write them, because you know your business and your customers. It suggests common questions and checks that each answer is short and direct, but every word is yours, and nothing appears on your site until you approve it.
Will the FAQ code get me a special result in Google?
Probably not. Google stopped showing FAQ rich results for most websites in 2023. The FAQ code is there so search engines and AI assistants can read each question and its answer clearly.
Can I use it with Yoast SEO or Rank Math?
Yes. If your SEO plugin already adds FAQ code to the same page, for example through its own FAQ block, use one or the other on that page, not both.
Why does it check Googlebot?
As a control. If your site refuses AI crawlers but serves Googlebot, that points at an AI-specific rule. If it refuses both, you have a general bot problem that is hurting your Google ranking as well, which is more urgent.
Screenshots



Changelog
1.1.0
- New Questions & answers tab. Write the questions customers ask with short answers, approve them, and show them anywhere with the [foundash_qa] shortcode, with FAQ code built from the same text. Suggested questions are filled in from your business and town. Answers are checked for being missing, too long or slow to get to the point. Nothing is shown until you approve it.
- OpenAI’s search crawler and training crawler are now requested with the user agent strings OpenAI currently publishes (OAI-SearchBot 1.4, GPTBot 1.4).
- Deleting the plugin now removes everything it stored.
1.0.7
- One line under the results now says what the plugin cannot tell you (whether ChatGPT names your business) and links to Foundash, where the first check of that is free. The link is plain, with no tracking. No change to the checks.
1.0.6
- Links now point to getfoundash.com, the new home of Foundash. The link to the 293-site research now goes straight to it. No code changes.
1.0.5
- If robots.txt cannot be read (the server errors, rate-limits the request or does not answer), the check now says it could not confirm access. Before, it treated an unreadable robots.txt as one with no rules and could report that AI assistants can read your site when it had not checked. A missing robots.txt (404) still means nothing is blocked.
- Crawlers that are only checked in robots.txt now show “Could not test” when robots.txt could not be read, not “Allowed”.
- The description now says which agents are requested live and which are only checked in robots.txt.
1.0.4
- New name and tags so the plugin is easier to find when you search for an AI crawler check. No other changes beyond 1.0.3.
1.0.3
- Fixes the “Discourage search engines” setting from this screen instead of telling you where to find it. The check already read the option directly to name the cause; now there is a button that turns it off, and a link to Settings > Reading for anyone who would rather do it themselves. It is reversible from the same place.
- The version constant had drifted from the plugin header (1.0.1 against 1.0.2), so an upgrade did not clear the cached result it was supposed to clear. Both now read 1.0.3.
1.0.2
- Links to the research the plugin is based on: 293 local business sites measured for AI crawler blocking.
1.0.1
- Separates the agents that read your page when a customer asks from the crawlers that collect pages to train on, and only raises an alarm for the first kind. Adds ChatGPT-User as a request and Perplexity-User and Claude-SearchBot as robots.txt checks. Version 1.0.0 reported a training opt-out as “ChatGPT: blocked”, which reads as “you are invisible” when the site is still perfectly readable at the moment a customer asks about it. That was measured, not guessed: of 293 local business sites, 14 of the 15 refusing an AI crawler in robots.txt refused only the training crawlers and allowed the search-time agents by name.
1.0.0
- First release. Crawler access checks for ChatGPT, ChatGPT Search, Claude and Perplexity using each one’s real user agent, robots.txt parsing with correct group precedence, detection of WordPress’s “Discourage search engines” setting, structured data detection, llms.txt detection, and a Googlebot control. robots.txt and llms.txt are read at the domain root, so a WordPress install in a subdirectory is checked against the file crawlers actually read. Structured data is read by decoding the JSON-LD rather than pattern matching it, so a type written as a list is counted. Runs are held to a wall clock budget and report themselves as incomplete rather than timing out. Detects a noindex tag on the front page, which is what WordPress’s “Discourage search engines” setting has actually set since WordPress 5.3, and what an SEO plugin may set on its own.
