Public status API

A status page is public information by design, so reading it is public too. These two endpoints need no authentication or key: they are the same data your public page itself renders.

Writing is a different matter: see the Incident API for creating and updating incidents. For a rendered badge or an RSS feed rather than JSON, see Badges & feeds.

Base URL

Your own page's hostname, not ours: your subdomain, or your custom domain once it is live:

https://status.example.com/api/public/{slug}

The slug appears twice, and it must be the slug belonging to that hostname. Asking one page's host for a different page's slug returns 404 rather than that page's data.

GET/summary

Current state and uptime history for every service on the page.

curl https://status.example.com/api/public/{slug}/summary

const response = await fetch('https://status.example.com/api/public/{slug}/summary');
if (!response.ok) throw new Error(`InBrief responded ${response.status}`);
const summary = await response.json();

{
  "slug": "acme",
  "lastChecked": "2026-08-20T21:02:09.028Z",
  "monitoringPaused": false,
  "services": [
    {
      "name": "Checkout API",
      "names": {
        "en": "Checkout API",
        "ar": "واجهة الدفع"
      },
      "url": "",
      "icon": "https://status.example.com/icons/checkout.svg",
      "slug": "checkout-api",
      "status": "up",
      "uptime": "99.98%",
      "uptimeDay": "100.00%",
      "uptimeWeek": "99.95%",
      "uptimeMonth": "99.98%",
      "uptimeYear": "99.98%",
      "time": 142,
      "timeDay": 138,
      "timeWeek": 141,
      "timeMonth": 142,
      "timeYear": 142,
      "dailyMinutesDown": {
        "2026-08-14": 7
      },
      "dailyHealth": {
        "2026-08-14": {
          "rate": 0.486
        }
      }
    }
  ]
}

How far back the daily maps go follows the page's own history window (7 to 90 days, see Branding).

?include=response-time

Ask for the daily response-time trend as well. Without it the response uses the default service fields; any other include value is a 400 rather than a silent fallback to the default.

curl https://status.example.com/api/public/{slug}/summary?include=response-time

const response = await fetch('https://status.example.com/api/public/{slug}/summary?include=response-time');
if (!response.ok) throw new Error(`InBrief responded ${response.status}`);
const summary = await response.json();

{
  "responseTime": {
    "windowDays": 30,
    "averageMs": 142,
    "points": [
      {
        "day": "2026-08-18",
        "averageMs": 138
      },
      {
        "day": "2026-08-19",
        "averageMs": 151
      }
    ]
  }
}

This object is added to each service. What it measures, exactly:

Only completed outbound checks count. A check inside an announced maintenance window is excluded, exactly as it is excluded from uptime; a cron heartbeat has no outbound duration to measure at all and always reports an empty series. A service with nothing eligible in the period returns "averageMs": null with no points, never 0, which on a latency figure would read as "instant" rather than "not measured". History begins when this representation was first published: earlier days were never recorded in a form that could be separated from maintenance, and they are left out rather than estimated.

GET/incidents

The incident ledger: open and resolved, incidents and scheduled maintenance alike.

curl https://status.example.com/api/public/{slug}/incidents

const response = await fetch('https://status.example.com/api/public/{slug}/incidents');
if (!response.ok) throw new Error(`InBrief responded ${response.status}`);
const ledger = await response.json();

{
  "slug": "acme",
  "incidents": [
    {
      "id": "inc_example_01",
      "type": "incident",
      "services": [
        "checkout-api"
      ],
      "startedAt": "2026-08-14T09:00:00.000Z",
      "resolvedAt": "2026-08-14T09:41:00.000Z",
      "publishedAt": "2026-08-14T09:00:10.000Z",
      "scheduledEndAt": null,
      "maintenanceState": null,
      "title": {
        "en": "Elevated error rates"
      },
      "summary": {
        "en": "We are investigating elevated error rates."
      },
      "updates": [
        {
          "id": "update_example_01",
          "status": "identified",
          "body": {
            "en": "A failover has been triggered."
          },
          "createdAt": "2026-08-14T09:12:00.000Z"
        }
      ]
    }
  ]
}

Caching

Both endpoints answer with cache-control: public, max-age=60. Poll no faster than that; a shorter interval gets you the same response from a cache, not fresher data.

The ?include=response-time representation is cached differently, because it is what the status page itself reads on a timer: cache-control: public, max-age=0, s-maxage=30, must-revalidate with a strong ETag. Send it back as If-None-Match and an unchanged page answers 304 with no body. The two representations are separate cache entries; asking for one never returns the other.

If you want push rather than poll, webhooks deliver these same events to your endpoint as they happen.

Errors

There is nothing else to handle. No authentication means no 401, and no plan gate means no 403.