status: 1 — quarantine. Yours will be too. This is
normal and it is not a restriction on your account.
What it actually does
A quarantined link runs a live safety check on the destination at each resolution. An active link (status: 0) does not.
That is the whole difference. A quarantined link resolves. Scanning the code
works. Nobody sees a warning, an interstitial, or a delay they would notice.
How a link leaves quarantine
It earns it:1
Three clean checks
Observed at resolution, on real traffic.
2
At least seven days old
Both conditions, not either.
That is deliberate, and it is load-bearing. If release depended on subscription
status, something on the redirect path would have to ask about billing — and
the guarantee in Permanence would become a policy we intend to
keep rather than a fact about the system.
Editing sends it back
Change a destination and the link returns to quarantine, its verification count resets to zero, and its version increments. This is not bureaucracy. Bank three clean checks on a harmless destination, wait for release, then swap in a malicious one, and you would have a link the fast path has stopped checking. The version guard closes it: every verification event carries the version it was observed under, so a check still in flight when you edit credits nothing toward the replacement. The cost to you is that a repointed code re-earns release from scratch. It keeps resolving throughout.Statuses
A link at
2 is never rewound by an edit or by a verification. That is the one
state that does stop a code resolving, and it is reserved for destinations that
failed a real safety check.
Destinations are checked at write time too
Every destination is validated and scanned before it is stored — on creation and on every edit.http/https only; no localhost, no private ranges, no cloud
metadata endpoints. A destination that fails returns 422.
Quarantine supplements that check rather than replacing it: the live re-check at
resolution fails open by design, because a safety-provider outage must never
break a printed code. So the write-time scan is what stands between a never-scanned
destination and being served during such an outage.