Run this domain's source classification now
Enqueues one domain's source-classification pass instead of waiting for the nightly run.
The readiness endpoint's impact is the figure the verdict turns on, and
it reports its own completeness. An agent that reads
completeness: partial and can do nothing about it until tomorrow gives
the customer worse advice than the dashboard does — this is the endpoint
that closes that gap.
One run per domain per hour, and a spent window is a 200. Being
early is not a client error: the answer is "the last pass was recent
enough", and the body says so with queued: false, an honest
retry_after_seconds, and the two figures a caller actually wants —
when the last pass ran and what it left unmeasured. Retry-After is set
on the response as well. The limit is a rule about a DOMAIN, not about a
caller: it is the same window and the same limiter the dashboard's
control uses, so the two surfaces cannot quote different waits.
The response describes the registry as it stood when the request arrived. It is never a prediction of what the queued run will find.
This endpoint writes no DNS, changes no DMARC policy, and records no decision about a source. It reruns a measurement over evidence already received. Non-destructive, so a test-environment credential may call it.
Path parameters
domain string required The domain, by NAME (acme.com). A UUID is also accepted. Matched case-insensitively, and scoped to the calling org — a domain belonging to another org is a 404, never a 403.
Responses
Errors follow the RFC 7807 problem format — see the error reference.
queued boolean required False when the window was already spent. Not an error — the answer is "the last one is recent enough".
retry_after_seconds integer required Seconds until the next run is allowed; zero when queued. Honest to the second: a refused call is not recorded against the window and so cannot push this out.
last_measured_at string required When the classification pass last wrote a row of this domain's registry. Empty when it never has.
unmeasured_sources integer<int64> required Registry rows carrying no measurement.
unmeasured_max_messages integer<int64> required The MOST those rows could add to impact.would_affect_messages: their whole thirty-day volume, assuming every message failed. An upper bound, never an estimate.
registry_sources integer<int64> required The size of the registry, so the counts read as a share.
code is internal_error. The
request_id field can be quoted to SendOps support to investigate.
application/problem+json