Checking whether your pages are actually indexed
You published a page. Is it actually in Google? Is Bing aware of the site at all? Did anything ever get told the page changed? A crawler cannot answer any of that by reading your HTML, because none of it is a property of your page — it is a property of somebody else's index.
Docket answers it by asking each engine directly, with your own credentials, and the interesting part of the answer is where it refuses.
What each engine can actually tell you
These are not equivalent sources dressed up as one number. Each one supports a different claim, and the row says which.
Google, per URL
The URL Inspection API returns a per-URL verdict and a last-crawl date for a property you have verified in Search Console. It is the only engine here that answers per page, and it can, because the published quota is 2,000 inspections per site per day and 600 per minute — enough for a whole site, every day.
Docket reads the verdict, which is a documented enumeration, and shows Google's own summary sentence beside it without interpreting it. That sentence is a plain string in the API, it has no published list of values, and the request takes a language code — so the same state arrives in different words for different accounts. Matching on the words would report an indexed page as missing the day Google rewrote them, or for anyone not working in English. Only the two decisive verdicts become a claim; the rest read as unknown rather than being guessed into a finding.
Bing, for the site — not for the page
Bing's crawl statistics give a count of pages in its index for the whole site, one row per day. There is no per-page column, and there is not going to be one: the per-URL call throttles after a few dozen requests, so a per-page Bing column would be blank or wrong on any site large enough to care. The row says site-level, and carries the date of the reading, because a count without its date is how a number goes stale while everything on screen agrees with it.
It is worth showing at all because Bing's index is not only Bing's. Several other search surfaces are served from it, which is also why Docket does not print the same figure again under their names.
Brave, for the site, with the ceiling printed
Brave can be enumerated by listing the host's results a page at a time, but its API stops at a fixed offset — about 200 results is the entire window that can ever be read for one host. Past that point, "not in the results" and "past the end of what I was allowed to read" are the same observation.
So beyond the window Docket reports blindness, not absence. It will never tell you a page is missing from Brave when what actually happened is that nobody could look. A tool whose whole promise is telling you where you are not indexed is exactly the wrong place to invent an absence.
IndexNow is your outbox, and it is labelled as one
IndexNow is how a site announces that a URL changed, to the endpoints its FAQ lists — Amazon, Bing, Naver, Seznam.cz, Yandex and Yep. Docket records whether a URL was announced and when.
That is all it records. Announcing is not indexing, and IndexNow's own FAQ says a submission "does not guarantee immediate indexing". In a table headed by the question "am I indexed", this is the single most misreadable column on the page, so it never counts toward the answer. The states it can report do not overlap with the index verdicts at all — it can say announced, or not announced, and nothing else.
AI crawlers, for the origin
Docket already checks whether your server actually serves the AI crawlers, by asking it — see AI crawlers and your sitemap and does noindex stop AI crawlers. That check probes your home page once per crawler, which makes it an answer about the origin.
It is joined to this table rather than rebuilt, and it is joined as what it is. It does not appear as a per-page verdict, because nothing about it was measured per page — and a column of crawler names and status codes looks far more specific than it is.
What this will not tell you
- Why Google has not indexed a page. It reports the verdict Google gives and shows Google's own summary; it does not diagnose a decision made inside somebody else's system.
- Per-page Bing or Brave state. Neither is available at a rate that would survive a real site, and a column that is right on small sites and wrong on large ones is worse than no column.
- Whether Yandex, Naver, Seznam or Yep indexed anything. Docket can announce to them and read nothing back. Announcements are shown; state is not invented.
- A single "indexed everywhere" score. Four sources answering four different questions do not average into one number, and a number that hides which engine said what is a number you cannot act on.
Where your URLs go, stated plainly
This is the one feature in Docket that talks to a search engine about your pages, so it should be obvious what that means.
Your URLs do go to Google and to Bing. That is not a side effect, it is the feature: to ask whether Google has your page, Docket has to tell Google which page you mean. The requests are made from your Mac, with credentials you supply, against properties you have verified as yours. Nothing is routed through us, nothing about your site reaches us, and there is no account with us to hold it. If you would rather not ask, the connector is a switch and the rest of the audit is unaffected.
The credentials are yours and stay on your machine. Docket never ships a shared key for this: a shared key would mean your URLs travelling under somebody else's identity, and a per-customer bill with no ceiling.
Where to start
The reads that need no credentials work immediately. Google needs a Search Console property you have verified; Bing needs its own key. You can run Docket against your site without either and still get the site-level picture, then add the sign-in when you want the per-page column. What the rest of the audit covers is on what Docket checks.
Sources read on 16 September 2026: Google's URL Inspection reference and published quota, Microsoft's crawl-statistics reference, and IndexNow's FAQ.
Common questions
Can Docket tell me if a specific page is in Bing?
No, and no tool can do it reliably at scale. Bing's per-URL endpoint throttles after a few dozen calls, so a per-page column would be blank or wrong on any site big enough to need one. Docket shows Bing's site-level count with the date of the reading instead, and says that is what it is.
Does Docket send my URLs to Google?
Yes, for the pages you ask it to inspect. There is no way to ask whether Google has a page without naming the page. The request goes from your Mac using your own Search Console sign-in against a property you verified, nothing is routed through us, and turning the connector off stops it entirely.
Why does a page say unknown instead of missing?
Because nobody looked. Brave's API stops after a fixed number of results per host, and past that point an absence from the results is not evidence of an absence from the index. Reporting it as missing would invent a finding on every page past the cutoff of every large site.
Is submitting a URL through IndexNow the same as being indexed?
No. IndexNow's own FAQ says a submission does not guarantee indexing. It records that you announced a change, which is a real and useful action, but it is your outbox rather than anyone's index. Docket labels it as a submission and never counts it toward whether a page is indexed.
Do I need an account with Docket for this?
No. Docket has no accounts. This feature uses credentials you already have with Google and Bing, entered on your own machine, and the reads that need no credentials work without any sign-in at all.
Which engines receive an IndexNow submission?
IndexNow's FAQ lists Amazon, Bing, Naver, Seznam.cz, Yandex and Yep as its endpoints. Docket does not keep its own copy of that list, because a list hardcoded in a tool goes quietly out of date while the tool keeps citing it.