Crawlune Docs
Reference

What leaves your site

The audit sends nothing at all. Only rewriting does, and only for the items in a run you started.

The audit is entirely local

Scoring your pages, checking your crawler controls and generating llms.txt all happen on your own server. Nothing is transmitted, and no licence is involved. That is the whole free half of the product.

When you rewrite content

Rewriting will happen on our service, so the text being rewritten is sent to it — the passage in question and enough of the surrounding page for the result to be consistent with it, plus your site's name and language.

Future tense deliberately: rewriting is switched off and a run currently sends a batch that comes back empty. This section is here so the answer is on record before the first byte of anybody's prose moves, rather than written afterwards to match what was built.

Not sent: pages that are not part of the run, your customers' data, user accounts, or content from anywhere else on your site.

The optional weekly report

Off unless you switch it on in Settings. Nothing prompts you to, and a site that never turns it on never sends anything.

What it contains, in full:

  • how many AI crawlers this site blocks, and how many of those are retrieval crawlers — counted, never named. Which crawlers you chose to block is your editorial decision and is not sent;
  • whether WordPress's own Discourage search engines setting is on;
  • whether an llms.txt is published;
  • once you have run an audit: how many pages were analysed, the median score, and the component counts behind it — how many pages are thin, lack an answer block, cite nothing external, contain no statistics. Per site, never per page;
  • your WordPress, PHP and Crawlune versions, your site's locale, and whether WordPress considers the site production, staging or local;
  • a one-way hash identifying this installation, salted with a random value generated on your own site and never sent. It lets one site's weekly reports be counted as one site; it cannot be turned back into an address.

Never sent: your site address, any e-mail address, any user account, any page content, any URL of yours, which specific crawlers you block, and your licence key.

A site that has never run an audit sends no scores at all rather than zeroes — it is unmeasured, not badly structured, and reporting the two the same way would corrupt the only number this is for.

Turning the setting off stops the weekly job immediately. Uninstalling deletes the salt, so a later reinstall is a different installation as far as we can tell.

Why we ask: how many real sites block a retrieval crawler decides whether that screen is the front door of this product, and the median score in the wild is the only honest basis for saying what a typical site's gap is. Both are currently guesses.

A note on this page's history. It previously described this report — in detail, including how a site would be identified — before any such thing was built. That was corrected to say so plainly, and the feature has since been built to match. The order matters: documentation describing collection that does not happen is the same defect as documentation omitting collection that does.

Uninstalling

Deleting the plugin removes its tables, options and scheduled jobs, and releases the site slot on your licence. Content it rewrote and you applied stays as it is — it is your content, in your posts. What goes is the record that would have let you undo it.

Crawlune documentation — pre-release