hosting
GET /v1/sites/{siteId}/index-check
A hosted site's search-index status.
Authenticatie
Stuur een API-sleutel mee als bearer token. Dit eindpunt vermeldt geen specifieke toestemming in de specificatie, dus geef uw sleutel de minimale rechten die nodig zijn en controleer het antwoord in plaats van zomaar wat aan te nemen.
Dit eindpunt vereist geen organisatie-id. Uw sleutel identificeert al de organisatie waartoe deze behoort, en het antwoord is hierop afgestemd.
Probeer het
Vervang alles tussen punthaakjes door uw eigen waarden en de sleutelplaatsvervanger door een sleutel uit uw dashboard.
curl -X GET https://api.zinndigital.com/v1/sites/{siteId}/index-check \
-H "Authorization: Bearer zdk_live_…"Ingelogd? De API-console in je dashboard vult je echte organisatie-id en je eigen sleutel in, en voert het verzoek uit tegen de live API zodat je de daadwerkelijke respons kunt zien. Open dit eindpunt in de API-console
Details
The **cached** search-index reading for this site's primary domain, in the same shape as the domain endpoint — it is the same underlying reading, reached from the site. ⛔⛔ **This is the only door most customers have.** The reading is keyed by FQDN and a hosted site has one, so it was always expressible; until this endpoint existed the only way to reach it was `GET /v1/domains/{domainId}/index-check`, which requires a `Domain` row. A Footprint-Free customer who registers their names elsewhere and connects nothing has no such row, so the status of their sites was unreachable from anywhere in the product. Never calls the upstream provider, so it is free to render. `state` distinguishes the same four cases the domain endpoint documents, and `unknown` must not be rendered like `not_indexed` — a client that conflates them reports healthy sites as deindexed during any provider outage. ⛔ There is deliberately **no** run counterpart on a site. Each check spends provider allowance, and a refresh button on every site across a large estate is an unbounded cost; freshness comes from the scheduled sweep. To force a check for one specific name, use the domain research tool, where the spend is metered against your own allowance.
Parameters
| Naam | Type | Verplicht | Wat dit is |
|---|---|---|---|
siteId (path) | Uuid | Ja | Site ID (UUIDv7). |
Reactie
| Naam | Type | Verplicht | Wat dit is |
|---|---|---|---|
fqdn | string | Ja | — |
state | string<indexed, not_indexed, unknown, never_checked> | Ja | `indexed` — the provider answered and the search engine has this site. `not_indexed` — the provider answered and the search engine has nothing. A real measurement, and the one s… |
checked_at | string | Nee | When the stored check ran. Null when never checked. |
is_stale | boolean | Ja | True when the reading is older than the 7-day monitoring window, or absent entirely. A hint that a fresh check is worth its allowance — not an indication the result is wrong. |
estimated_indexed_pages | integer | Nee | The search engine's own loose upper bound for how many pages of this site it holds. **An estimate, not an authoritative count** — it routinely differs from Search Console by sev… |
sample_urls | string[] | Nee | A few URLs the search engine returned, so a customer can confirm the right site was matched. Evidence, not a sitemap — capped at five. |
next_check_at | string | Nee | When this domain is next due for an automatic check. Indexed sites are re-checked monthly, missing ones daily, and unknown ones within hours. |
Fouten die dit eindpunt kan retourneren
401 · 403 · 404