links
GET /v1/link-targets
The Link Tracker report — domains this org's sites link out to.
Autentikasi
Kirimkan kunci API sebagai token bearer. Titik akhir ini tidak menyatakan izin tertentu dalam spesifikasi, jadi berikan kunci Anda izin minimum yang diperlukan dan periksa responsnya alih-alih berasumsi.
Tempat ID organisasi Anda dimasukkan
Titik akhir ini menerima org_id sebagai parameter kueri. Kosongkan parameter tersebut dan panggilan ini akan mencakup seluruh subpohon penyewa Anda; kirimkan untuk mempersempit panggilan ke satu organisasi.
ID organisasi Anda berada di layar kunci API di dasbor Anda, di sebelah kunci itu sendiri. Itu adalah ID yang sama dalam setiap panggilan yang Anda buat.
Coba
Ganti apa pun di dalam tanda kurung sudut dengan nilai Anda sendiri, dan placeholder kunci dengan kunci dari dasbor Anda.
curl -X GET https://api.zinndigital.com/v1/link-targets \
-H "Authorization: Bearer zdk_live_…"Sudah masuk? Konsol API di dasbor Anda akan mengisi ID organisasi asli dan kunci Anda sendiri, serta menjalankan permintaan terhadap API langsung sehingga Anda dapat melihat respons aktualnya. Buka titik akhir ini di konsol API
Detail
V1's columns, rebuilt: **Domain | Site | Reflinks | Ref posts | Ref sites**. Ranked by reflinks by default, because the question is *"who am I linking to most"*. **Search rolls variants up.** `www.`, letter case and punycode/Unicode are folded on the **write** path, so they are already one row by the time you search — that is the fix for the production defect where `cu-tv.com` showed 185 reflinks and `www.cu-tv.com` showed 1 (docs/96 §1.4, D1328). **Subdomains** stay separate rows, because an operator cares which host they hit, but a search for `skokka.com` finds `blog.skokka.com` too and `rollup` gives the combined totals across the whole match. ⛔ **Paged by `offset`, not by cursor.** The shared cursor paginator is keyset over the sortable id, which is chronological; this list is ranked, and a keyset cursor cannot express that ordering. `total` is the full match so the screen can page all of it — there is no hidden cap. `coverage` says how much of the estate the report is based on. ⛔ Without it a customer cannot tell *"no outbound links"* from *"nobody has scanned this yet"*, and those two readings lead to opposite actions.
Parameter
| Nama | Jenis | Wajib | Tentang apa ini |
|---|---|---|---|
org_id (query) | string | Tidak | Narrow to one organisation in the caller's subtree — what the dashboard's org switcher sends. ⛔ **Narrow only:** it is intersected with what the caller may already read, so an o… |
q (query) | string | Tidak | A domain, or any text. A value that parses as a hostname matches that host **and every subdomain of it**; anything else is a substring match, because someone typing `casino` mea… |
owned (query) | string<own, external> | Tidak | `own` shows only targets that are the customer's **own** sites; `external` only those that are not. Omitted shows both. |
site_id (query) | string | Tidak | Only targets linked to **from** this site. |
include_ips (query) | boolean | Tidak | Bare-IP targets are real links and are tracked, but they never group sensibly. Set `false` to hide them. |
min_links (query) | integer | Tidak | The noise filter. Padding links (`google.com`, `youtube.com`) sit beside the money links deliberately. ⛔ They are **filterable, never auto-deleted** — removing a link a customer… |
sort (query) | string<links, posts, sites, host, recent> | Tidak | — |
limit (query) | integer | Tidak | — |
offset (query) | integer | Tidak | — |
Tanggapan
| Nama | Jenis | Wajib | Tentang apa ini |
|---|---|---|---|
items | LinkTarget[] | Ya | — |
total | integer | Ya | — |
offset | integer | Ya | — |
limit | integer | Ya | — |
has_more | boolean | Ya | — |
rollup | LinkRollup | Ya | Combined totals across **every** target the filters matched — the *"including variants of it"* number. Not the sum of the page, so it does not change as you page. |
coverage | LinkCoverage | Ya | How much of the estate the report is based on. ⛔ Without it a customer cannot tell *"no outbound links"* from *"never scanned"*, and those lead to opposite actions. |
Kesalahan yang dapat dikembalikan oleh titik akhir ini
401 · 403 · 429