Service status

We run a public uptime page that polls the real health endpoints of the platform every 5 minutes. The code that generates it lives at Neuroon-ai/status and the rendered page is served at:

https://status.neuroon.ai

Monitored components

Component	What it checks
API HTTP (production)	api.neuroon.ai/actuator/health — Spring Actuator returns {"status":"UP"} when the API, database and critical dependencies are healthy.
API HTTP (development)	dev-api.neuroon.ai/actuator/health — staging environment.
Widget CDN	cdn.neuroon.ai/versions.json — manifest of published versions. A 200 means the CDN is serving.
Documentation	docs.neuroon.ai/ — the docs site (Firebase Hosting).
Dashboard	neuroon.ai/dashboard — admin UI.

How it works

The monitor uses Upptime, open-source. Each check:

1. A GitHub Actions cron runs */5 * * * * and GETs every endpoint.
2. If the response is not the expected one (status code or latency over threshold), an issue is opened in the Neuroon-ai/status repo and the team is notified.
3. When the component is back, the issue is closed automatically.
4. Every 8 h the static page is regenerated with the 90-day uptime history.

Data is not hand-edited: each history/*.yml in the repo is the direct evidence of each check, with the returned status code and response time.

SLA policy

• We do not publish a contractual SLA with an uptime percentage outside Enterprise contracts. We run continuous observability and prioritize incident resolution, but we do not commit numbers like 99.9% or 99.95% by default.
• The Enterprise plan includes a negotiable SLA. Talk to your account contact if you need one.
• The historical uptime shown on the status page is real data, not a marketing claim. If a component goes down, the page reflects it.

Reporting an incident

When you spot something that looks like a failure:

1. Check https://status.neuroon.ai first. If there is an open incident matching your symptoms, wait for the official update.

2. If there is no public incident, open a ticket with Support including:

   • Environment: Production or Development.
   • Shop ID.
   • request id if your HTTP client logs it (header X-Request-ID).
   • timestamp from the error response body.
   • Endpoint (e.g. POST /api/plugin/shops/.../products/sync).
   • Anonymized payload (no PII, no keys).
   • Exact response, including X-RateLimit-* headers if they apply.

3. For incidents of broad impact (API outage, extreme latency) we prioritize response during European business hours. Enterprise customers have escalation routes outside business hours.

Sensitive data in tickets

Do not send API keys or personal data in tickets. If we need to correlate, we'll ask you for a shop id or request id and cross it with our logs.

Subscribe to updates

• GitHub Watch: https://github.com/Neuroon-ai/status — click Watch → Custom → Issues. Every open incident gets emailed to you.
• RSS: the status page exposes an RSS feed with every status change.

Scheduled maintenance

• We give at least 48 h of notice.
• Posted at https://status.neuroon.ai and emailed to account administrators.
• We schedule during low-traffic windows (European nights) when possible.

Further reading

• Support — channels and response times.
• Production — availability policy.
• Changelog — change history.