The questions people actually ask.

Enter a domain. We crawl up to 10 pages the way Googlebot and the AI crawlers do, run all 129 checks, and show the score, grade, category breakdown, coverage, and how many findings there are. No signup, no card.

A crawl costs money and the free scan is the front door, not the product. It always reports how many pages it covered against how many it found.

The score is the diagnosis and it is free. The findings are what an email buys. The fixes are what a plan buys. A shared scorecard link never unlocks the findings, whoever opens it, and expires after 30 days.

One real crawl per domain per day. A repeat request returns the scan that already ran, which also stops the endpoint being used to point traffic at someone else's site.

The check could not run: the page was unreachable, a resource timed out, an external source was down. It is not counted as a pass. The industry default is to round it up, which is how tools report 95/100 for sites with real problems.

Usually because they counted what they could not check as fine. Compare coverage, not scores. A 94 at 40% coverage and a 71 at 95% coverage are not describing the same thing.

SEO checks cover what search engines rank: metadata, structure, speed, links. The 16 GEO checks cover what answer engines cite: whether ChatGPT, Claude, Perplexity and AI Overviews can lift your content and attribute it to you.

For some checks, yes, and those are labelled Model wherever they appear. Most of the rubric is rules and graph analysis. A model judgement is never presented as a fact.

The loop stops before any fix is written, where you review the ranked plan, and before the finished fixes are handed to you, where you mark them ready. Crawld writes; you decide and you apply. Nothing reaches your site except by your hand.

Written out in full: the exact text or tag, where it goes, what it replaces, in language you can paste into your platform's editor. Connectors that write drafts into your CMS are in progress; the integrations pages track each one.

Every fix shows the exact before and after before you apply it. Applied changes are re-measured on the next scan; if one did not move its check from fail to pass, it is flagged. Because you apply each change yourself, undoing one is the same edit in reverse.

The editor drafts against the gaps the audit finds and checks every draft against what you have already published so your posts do not compete with each other. Claims about your product and whether a page should exist stay with you. It is in Pro.

Nothing today. Every fix is advisory: written out, applied by you. If a repository mode ships, that badge will appear only there.

Never. There is no code in Crawld that can log in, comment, vote or message on Reddit. You copy the draft, edit it, and post from your own account. That is why the drafts are worth posting.

No. Drafts are yours to copy into your CMS. When you have published one, paste its URL and the next scan measures it.

The scan reads what a crawler reads. If your site renders client-side, many checks will show as unmeasured, which is the finding. The fix queue tells you what to change in your builder.

The free scan, which needs no access, and the scan API. Every fix arrives written out for you to apply. Connected delivery modes are planned, not built, and each integration page marks its own status.

We store the crawl results and the fixes we write for you. We do not resell crawl data and do not train models on your content. Any access you grant later is limited to the scopes you choose, revocable any time, and the data goes with it.

It identifies itself as CrawldBot and the user agent links to a page explaining what it does and how to block it. It never crawls under a browser user agent.

Disallow CrawldBot in robots.txt and it stops on the next run. Monitoring is capped at 12 pages per site, one request at a time, once a week.