Methodology
Methodology
The free check fetches a public URL and scores crawler access and the static HTML. It does not call a model. Mentions are not implemented in this version and have no score weight.
Free check
- We parse robots.txt and separate search-crawler rules from training-crawler rules.
- We request the homepage with those User-Agent strings and record the HTTP status.
- We record noindex, nosnippet, title, meta description, sitemap.xml, and JSON-LD types.
- Organization and SoftwareApplication are the JSON-LD types that satisfy that item. FAQPage and any other type are named and do not replace them.
- Text length is the static HTML. JavaScript is not executed.
- llms.txt is recorded with weight zero.
- About 3 checks per IP per day. A domain may be cached for 24 hours.
- The result is stored at /r/ and a random id. Anyone with the link can open it. It is not listed.
What is scored now
The free result shows the crawl items below. A model visibility score is not part of this version and is not for sale. Mentions are not tested and are not scored.
| Item | In the score | What we record |
|---|---|---|
| robots.txt could be fetched | Counted | A readable robots.txt response |
| Search-crawler rules | Counted | Search crawlers are not disallowed for the page |
| Live search-crawler requests | Counted | The homepage responds, and is not 401, 403, or 5xx |
| Training-crawler rules and requests | None | Recorded only. A block is not a failed check |
| noindex | Counted | The page does not send noindex |
| nosnippet | Counted | The page does not send nosnippet |
| Static text | Counted | The HTML has readable text without JavaScript |
| Title, meta description, sitemap, Organization and SoftwareApplication JSON-LD | Counted | Each one that is present |
| llms.txt | None | Noted only |
| Mentions and citations | None | Not tested in this version |
$19 report
The same crawler, meta, and structure evidence on pages we fetch from the homepage, the sitemap, and same-origin links. Up to 10 pages can be inspected. The cap is 10 pages. That is a cap, not a promise that 10 pages exist. One qualifying page is listed as one page. A priority is included only with the page URL and a verbatim excerpt from that fetch. Checkout is not open.
$29 pack
The report, plus files built from those fetches:
- robots.txt.diff from the robots.txt we fetched.
- An Organization and SoftwareApplication candidate from the observed title, description, and URL.
- Priorities that cite the source URL and excerpt.
- Question text copied from a fetched page only when that question is already there.
FAQPage markup, comparison outlines, and an llms.txt file are not part of this pack. Nothing is emailed. The customer page is /access/ plus a random token. That page reads the private record at /api/access/:token. It is not /r/ and not /private/. GET /api/orders/:id does not return the token.
What we do not do
- Call a model, or report mentions or citations
- Send email
- Change your site or contact other sites
- Treat a training-crawler block as a failed check