What we commit to on uptime and incident response for partner studios — set at numbers we can actually hit, not numbers that sound good in a deck.
This Service Level Agreement ("SLA") supplements the Terms of Service and applies to partner studios on paid tiers (Cerberus or Aegis) integrating the Cerberus Anti-Cheat SDK and API. It defines our uptime commitment, service credit schedule, and incident response targets.
We commit to 97% monthly uptime across the combined Detection Engine, API & Dashboard, and Data Pipeline services, measured across all three of our operating regions (US-East, EU-West, AP-Southeast).
This number is set deliberately at or below what we already deliver. Our public Status page shows actual trailing uptime in the 97–98% range — this SLA formalizes that as a commitment rather than promising a figure we haven't demonstrated.
Uptime is calculated per calendar month as:
(total minutes in month − downtime minutes) / total minutes in month × 100
Downtime means the covered services are unreachable or returning failure responses for the majority of requests in a given minute, as recorded by our monitoring and reflected on /status/. Scheduled maintenance windows announced at least 48 hours in advance via the status page are excluded from downtime calculations, as are the exclusions listed below.
If we miss the 97% monthly uptime commitment, partner studios on paid tiers are eligible for a service credit against that month's fees, applied to the following month's invoice.
| Monthly uptime | Service credit |
|---|---|
| 97.0% or higher | No credit — target met |
| 95.0% – 96.99% | 5% of monthly fees |
| 90.0% – 94.99% | 10% of monthly fees |
| Below 90.0% | 15% of monthly fees |
Total service credits in any single month are capped at 15% of that month's fees. Service credits are your sole and exclusive remedy for a missed uptime commitment — this section works alongside, not in place of, the liability terms in our Terms of Service.
Incidents on /status/ are tracked through the same lifecycle you can watch in real time: Investigating → Identified → Monitoring → Resolving → Resolved. We commit to the following response targets by severity:
These are closed-beta targets, not sub-minute enterprise SLAs — we're a 32-person team and we'd rather set targets we hit reliably than ones we quietly miss. Issues already disclosed under Known Issues on the status page are tracked there and are not newly eligible for these response commitments until they're reclassified as an active incident.
This SLA does not apply to unavailability caused by or resulting from:
Service credits aren't automatic during beta — request them through the Billing section of your partner dashboard, or email cerberus@arsenalrx.dev within 30 days of the end of the affected month. Reference the relevant incident ID from /status/ where applicable. We'll confirm eligibility and apply approved credits to your next invoice.
We may update this SLA as Cerberus moves from closed beta toward general availability. Material changes — including any change to the uptime commitment or credit schedule — will be communicated to partner studios via email and posted on our changelog at least 30 days before taking effect.
Questions about this SLA or a specific incident? Reach out:
General / partner support: cerberus@arsenalrx.dev
For the underlying legal terms, see our Terms of Service, Privacy Policy, and Data Processing Agreement. For our full security posture, see the Trust Center.