Define and publish a real SLA and incident-response commitment

[JSONbored]

Problem: Today's support posture is pure open-source ("use GitHub issues") with no SLA or incident-response time commitment — appropriate for a self-host OSS project, not for a paid hosted product.

Area: ORB / Incident readiness

Proposal: Define and publish a real SLA and incident-response process specifically for the hosted product, distinct from the existing OSS support posture.

Deliverables:

• A published SLA and incident-response commitment for hosted customers.

AMS Cloud Readiness cross-reference (added 2026-07-12)

See also #5227 in the new AMS Cloud Readiness milestone (#28) — the AMS-side parallel to this exact problem, already citing this issue as precedent. Keep both sides in sync if either design changes.

[linear-code]

JSO-3595

[JSONbored]

Superseded — converged into #5227's design, which explicitly covers "hosted AMS + ORB" (one shared SLA/incident-response commitment rather than two divergent ones, per #7173's shared-control-plane direction). See #5227 for the settled draft: 99.5% uptime, next-business-day response, single-operator escalation path today, marked as a starting point pending real operational data.