Run now

After you fix an outage you do not have to wait for the next period. "Run now" runs a check once, right away, in every region.

From the dashboard

Open the check and click Run now in the page header. The button shows a spinner until a fresh result has arrived for each region, then a toast shows the new status. If nothing arrives within 2 minutes it tells you it is still waiting. Viewers do not see the button.

From the API

curl -X POST -H "Authorization: Bearer $TOKEN" \
  https://your-host/api/v1/orgs/default/checks/$CHECK_UID/run-now

{
  "requestedAt": "2026-10-04T10:15:00.123456Z",
  "regions": [
    { "region": "eu", "status": "queued" },
    { "region": "us", "status": "running" }
  ]
}

• queued: the region's run starts now.
• running: the region is already running the check (or a multi-step run is in progress). That result answers your request.

A result with a periodStart at or after requestedAt is the answer.

What it does to incidents

The result is a normal result. It can open or resolve an incident like any scheduled one, and confirmation and recovery periods still apply.

The next scheduled run keeps its regular period.

Limits

Check type	Per check	Per organization
Most types	3 per minute	60 per hour
browser, js, rdp, vnc	1 per minute	20 per hour

A request that queues nothing (every region is already running) answers 200 and spends no budget.

A refused request answers 429 with a Retry-After header. A disabled check answers 409.

From MCP

The run_check tool takes a check UID or slug and returns the same response. It needs the mcp scope (not mcp:read).

A newly created check already runs at once, with no request needed.