Tonkean experienced a major incident on June 19, 2026 affecting Workflows Runtime, lasting 1m. The incident has been resolved; the full update timeline is below.
Affected components
Update timeline
- identified Jun 19, 2026, 09:07 AM UTC
We're seeing an issue with agents not responding at all at the moment. We've identified the root cause, and are working on a fix.
- resolved Jun 19, 2026, 09:09 AM UTC
Incident is resolve, AI Agents are back working and responding.
- postmortem Jun 23, 2026, 02:41 PM UTC
On 19 June 2026, between 08:30 and 09:08 UTC, all AI agents were unavailable across Tonkean’s production environments for approximately 38 minutes. Agents did not respond to requests during this window, interrupting active work and demonstrations. We recognize the significance of this disruption and appreciate your patience. The outage was caused by Tonkean’s AI provider account reaching its usage and spending limits, which resulted in every agent request being rejected. Two factors extended the impact: the account’s quota had been consumed more quickly than anticipated, and an outstanding account balance had to be cleared before the quota could be restored — even after we had raised the limit. An automated alert had fired the previous afternoon, warning of the approaching limit, but it posted silently and was not acted on. Service was restored at 09:08 UTC after the balance was cleared and the limit raised. No data was lost. To prevent recurrence, we are wiring the existing alert to actively page the on-call team and billing account owners rather than posting silently, and adding a proactive alert at roughly 90% of the spend limit. We remain committed to the reliability of the Tonkean platform and to ensuring the gaps identified here are fully closed.