All systems operational

No incidents reported.

Before public beta, this page shows a sample status view and availability model to demonstrate how SkyAIApp will publish monitors, incident timelines, and subscriptions. Live probe data will replace it once production monitoring is public.

Last updated · 16 monitors · 16 operational

Status subscription channel preview; live monitoring subscriptions open with public beta.
Sample 90-day uptime
99.414%
Monitors up
16/16
30-day avg response
173ms
Incidents (90d)
3

Core Platform

Last 90 days
API GatewayOperational
99.41% uptime174ms response
90 days agoToday
24h 99.82%7d 99.84%30d 99.86%90d 99.407%
Model RouterOperational
99.25% uptime186ms response
90 days agoToday
24h 99.93%7d 99.84%30d 98.88%90d 99.251%
Agent RuntimeOperational
99.64% uptime168ms response
90 days agoToday
24h 99.87%7d 99.86%30d 99.80%90d 99.643%
Semantic CacheOperational
99.74% uptime166ms response
90 days agoToday
24h 99.91%7d 99.88%30d 99.70%90d 99.738%
Policies & BudgetsOperational
99.50% uptime160ms response
90 days agoToday
24h 99.75%7d 99.88%30d 99.34%90d 99.500%

AI Provider Connectivity

Last 90 days
OpenAI (GPT-5.5)Operational
99.43% uptime195ms response
90 days agoToday
24h 99.99%7d 99.13%30d 99.46%90d 99.426%
Anthropic (Claude 4.7)Operational
99.31% uptime196ms response
90 days agoToday
24h 99.74%7d 99.85%30d 99.62%90d 99.313%
Google (Gemini 3.1)Operational
98.98% uptime173ms response
90 days agoToday
24h 99.78%7d 97.10%30d 99.23%90d 98.975%
DeepSeek (V4)Operational
99.36% uptime167ms response
90 days agoToday
24h 99.96%7d 99.85%30d 99.55%90d 99.363%
Mistral (3 series)Operational
99.59% uptime164ms response
90 days agoToday
24h 99.90%7d 99.81%30d 99.62%90d 99.587%
Meta (Llama 4)Operational
99.34% uptime164ms response
90 days agoToday
24h 99.75%7d 99.81%30d 99.72%90d 99.336%

Edge & Operations

Last 90 days
Console (app.skyaiapp.com)Operational
99.73% uptime168ms response
90 days agoToday
24h 99.92%7d 99.85%30d 99.67%90d 99.728%
Docs (skyaiapp.com/docs)Operational
99.09% uptime189ms response
90 days agoToday
24h 99.90%7d 96.92%30d 99.18%90d 99.088%
Marketing siteOperational
99.46% uptime157ms response
90 days agoToday
24h 99.95%7d 99.88%30d 99.86%90d 99.455%
Webhooks DeliveryOperational
99.16% uptime174ms response
90 days agoToday
24h 99.77%7d 99.83%30d 99.30%90d 99.163%
Billing & InvoicingOperational
99.64% uptime159ms response
90 days agoToday
24h 99.72%7d 99.87%30d 99.64%90d 99.643%
LegendOperationalDegradedOutageMaintenanceNo data

Sample incident timeline

Full changelog
MinorResolved

Elevated latency on Anthropic Claude routes (eu-west)

Model Router · Anthropic (Claude 4.7)
Started Apr 29, 02:12 PM UTC
Duration 1h 36m
  1. Resolved·

    Upstream provider recovered, p95 latency returned to baseline. Auto-fallback shielded 97% of traffic during the incident.

  2. Monitoring·

    Routes shifted to claude-haiku-4.5 and claude-opus-4.7 us-east; investigating regional issue with provider.

  3. Identified·

    Detected p95 latency >3s on Claude routes in eu-west. Triggering auto-fallback policy.

MaintenanceCompleted

Scheduled maintenance — Router v3 rollout

Model Router
Started Apr 15, 02:00 AM UTC
Duration 38m
  1. Completed·

    Sample Router v3 rollout completed with no service interruption in the modeled maintenance window. Regional fallback chains are active in the plan.

  2. In progress·

    Beginning rolling deployment of Router v3 with regional fallback. Traffic served by previous version during cutover.

MinorResolved

Webhook delivery delays

Webhooks Delivery
Started Mar 22, 09:05 AM UTC
Duration 1h 6m
  1. Resolved·

    Backlog drained, end-to-end delivery latency back to <2s. No webhook events were lost; affected events were retried.

  2. Investigating·

    Delivery latency rising past 30s on a subset of tenants. Investigating queue worker scaling.

MajorResolved

OpenAI upstream 5xx spikes (us-west)

Model Router · OpenAI (GPT-5.5)
Started Mar 4, 06:31 PM UTC
Duration 51m
  1. Resolved·

    Upstream stabilized. Automatic fallback to claude-sonnet-4.6 maintained 99.4% success rate during the spike.

  2. Identified·

    OpenAI status page reports a us-west incident. Fallback policy engaged for affected tenants.

Sample heartbeats use a 30s global edge-probe model; public beta will replace them with live probe data.Report an issue →
System Status | SkyAIApp — SkyAIApp