Speakeasy incident
Outage on MCP servers and all coding agent hooks
Speakeasy experienced a minor incident on September 9, 2026 affecting MCP Servers and Claude Hooks and 1 more component, lasting 12m. The incident has been resolved; the full update timeline is below.
Affected components
Update timeline
- investigating Sep 09, 2026, 03:30 PM UTC
Status: Investigating We are experiencing downtime on all services. Dashboard is currently inaccessible. The team is working on a resolution and we'll have an update soon. Affected components Dashboard (app.getgram.ai) (Full outage) MCP Servers (Partial outage) Claude Hooks (Partial outage) Codex Hooks (Partial outage) Cursor Hooks (Partial outage)
- monitoring Sep 09, 2026, 03:33 PM UTC
Status: Monitoring All systems are operational and the downtime is resolve. The Team is investigating the root cause and will post an update soon. Affected components Dashboard (app.getgram.ai) (Operational) MCP Servers (Operational) Claude Hooks (Operational) Codex Hooks (Operational) Cursor Hooks (Operational)
- resolved Sep 09, 2026, 03:43 PM UTC
Status: Resolved We have resolved the outage. A terraform bug: https://github.com/hashicorp/terraform-provider-time/issues/118 in the rotation logic for one of our client certificates caused the cluster to enter an unhealthy state. This was resolved through an internal fix to our terraform configs by setting up a new resource. We are monitoring the terraform bug and have reported upstream. We apologise for the inconvenience and do please reach out to our team on slack or [email protected] with any questions. We are here to assist. We will be adding additional validation in our infrastructure tests and monitors to catch similar bugs in cert rotation in the future. Affected components Dashboard (app.getgram.ai) (Operational) MCP Servers (Operational) Claude Hooks (Operational) Codex Hooks (Operational) Cursor Hooks (Operational)