Production incident support process

Q: Production incident support process

Check service status first. If there's a known incident, you don't need to open a duplicate ticket — the status page is the fastest source of truth. Classify the severity using support severity levels — SEV-1 for anything actually down in production. Collect safe diagnostic information before you write the ticket: the exact endpoint, the full response body with successful / message / data , the approximate time, and how many requests are affected. Never include your raw API key, a password, or an MFA code — see common API errors for what a useful report looks like. Open a ticket with Product/service set to Technical / API and severity set to match. For an active SEV-1, see escalating a critical issue for how to get it seen fastest.

Check status, classify severity, open a ticket with the right details.

Updated September 14, 2026 · 1 min read

Check service status first. If there's a known incident, you don't need to open a duplicate ticket — the status page is the fastest source of truth.
Classify the severity using support severity levels — SEV-1 for anything actually down in production.
Collect safe diagnostic information before you write the ticket: the exact endpoint, the full response body with successful/message/data, the approximate time, and how many requests are affected. Never include your raw API key, a password, or an MFA code — see common API errors for what a useful report looks like.
Open a ticket with Product/service set to Technical / API and severity set to match. For an active SEV-1, see escalating a critical issue for how to get it seen fastest.

Was this article helpful?

Still stuck?

Ask Kai about this article, or open a ticket with our team.

Submit a ticket