Act
Site health
Check whether AI crawlers can read your site, fix what blocks them, and prepare for Bing — the index behind ChatGPT search and Copilot.
Site health answers one question first: can the AI engines read your site at all? A site that blocks GPTBot or hides its content behind JavaScript gains nothing from any content work, so this audit is the first thing to run on a new workspace. The same screen also generates structured data and an llms.txt file, and walks you through Bing setup.
Where to find it
In the sidebar, open Site › Site Health. The page title is Audits and it has three tabs: Audit, Schema and llms.txt and Bing pack.
Run an audit
- Press Start audit in the page header.
- Enter the Address to audit — usually your home page, e.g.
https://yourbrand.com. - Press Start audit. The audit is queued and takes a few seconds to a couple of minutes.
- Reload the page to see the results.
The audit runs when you start it; run it again after each fix to confirm the change worked. A failing check also turns into an item in the action plan.
The seven checks
| Check | What it looks at | When it fails |
|---|---|---|
| robots.txt — AI crawlers | Whether robots.txt lets these crawlers in: GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, Perplexity-User, ClaudeBot, Claude-SearchBot, Google-Extended, Googlebot, Bingbot, CCBot, Meta-ExternalAgent. | One or two blocked is a warning; three or more is blocking. The finding lists the blocked crawlers and gives a robots.txt snippet to fix it. |
| Bot protection (CDN/WAF) | Rixe requests the page as a normal visitor and as an AI crawler and compares the answers. | The site answers the crawler differently (for example 403 instead of 200): allow AI crawlers in your CDN/WAF settings. If the site errors for everyone, fix availability first. |
| Content without JavaScript | How much of the page's text is in the raw HTML compared with the page rendered in a browser, and whether there is an <h1>. | Most of the content only appears after JavaScript runs. Most AI crawlers do not run JavaScript. Use server-side rendering or static generation. |
| Response speed (TTFB) | Time to first byte. | Over 800 ms is a warning, over 2,000 ms is blocking. Slow responses burn crawl budget. |
| Sitemap | Whether sitemap.xml exists and how fresh its lastmod is. | Missing is blocking; no lastmod, an unreadable date, or no update for more than 90 days is a warning. |
| llms.txt | Whether the site serves an llms.txt. | Missing is a warning: not required, but a cheap way to describe your site to AI crawlers. |
| Structured data (JSON-LD) | Schema types such as Organization or LocalBusiness on the page. | No structured data found. |
Content without JavaScript and Structured data can differ from page to page, so besides the address you entered, Rixe samples up to three more pages of different types from your sitemap and reports the worst result, with the page where it was found.
Read the results
Four tiles sum up the latest audit:
- Audit score — 0 to 100. Each check has a weight; a passing check keeps its full weight, a warning keeps half, a failure none. The crawler checks weigh most: robots.txt 30, bot protection 20, JavaScript 20, speed 10, sitemap 8, structured data 8, llms.txt 4.
- AI access — Open, Partial or Blocked, from the crawler-related checks. When blocked, the tile names the check responsible.
- Checks passed — for example "7 checks: 5 pass, 1 need fixing, 1 are blockers."
- Structured data — how many structured-data checks pass.
If a crawler is blocked, a red line at the top says Before anything else: and names it. When nothing blocks the crawlers, a green line confirms that AI engines can read your site.
The Technical access audit table lists every check with its status — blocking, fix this, pending or passing — sorted so blockers come first. Many findings include a copy-paste fix. On any failing or warning check, Open task creates a task for you or your developer in Tasks.
Let the plugin apply fixes
If your WordPress site is connected through the Rixe plugin, an Apply to the site card appears above the results with two fixes the plugin can make itself:
- Write llms.txt — publishes a map of your pages at the root of your domain.
- Add organisation data to every page — so machines no longer have to guess whose site this is.
Press Apply. The plugin picks the change up within the hour; to apply it at once, press Check for pending changes in the WordPress admin. Each fix shows queued, done or failed. WordPress keeps what was there before each change, so you can undo it from there with one click. See Integrations.
Schema and llms.txt generator
On the Schema and llms.txt tab, enter your brand name, site address and a short description. Optionally add a city (a local business schema is then generated) and one FAQ question and answer. Rixe produces:
- a JSON-LD block to paste into your pages'
<head>; - an llms.txt file to download with Download llms.txt and upload to your site root.
For an llms.txt built from your verified facts rather than a short form, use the one on your Profile — see Brand vault and knowledge base.
Bing pack
ChatGPT and Copilot draw answers from the Bing index, and few sites work on Bing at all — which makes it one of the cheapest visibility gains available. On the Bing pack tab:
- Enter your domain. Enter an IndexNow key or press Generate for a new one.
- Press Check steps. Rixe checks four steps and shows a Completion percentage:
- Bing Webmaster Tools verification — looks for the
msvalidate.01meta tag. The quickest way is to import your site from Google Search Console in Bing Webmaster Tools. - IndexNow key — looks for the key file at your site root. The instructions give the exact file name, address and contents.
- Submitting the sitemap to Bing — once verification is done, submit your sitemap in Webmaster Tools.
- Indexing in Bing — roughly how many of your pages Bing has indexed (this step needs the SEO data provider; without it the step says so).
- Bing Webmaster Tools verification — looks for the
- Follow the instruction under each step that is not done, then check again.
Plans
Site health — the audit, the plugin fixes, the generator and the Bing pack — is part of every plan.
Common questions
The audit says I block a crawler I never blocked. Check for a User-agent: * rule with Disallow: /, which blocks every crawler without its own group. The snippet in the finding adds explicit groups for the AI crawlers.
Bot protection fails but robots.txt passes. Your CDN or firewall (for example a bot-fight mode) is challenging crawlers before they reach robots.txt. Allow verified AI crawlers in its settings.
Where do I see which crawlers actually visit? In AI traffic and crawlers.
Related pages
- Action plan
- Site SEO — the full crawl of your site for classic SEO issues.
- AI traffic and crawlers
- Agent readiness
Still stuck? Write to [email protected] and a person will answer.