This describes how we respond to security incidents and serious outages.
1. Detect and triage
Incidents may be detected by monitoring, error alerts, user reports, or researcher reports to [email protected]. We assess severity based on whether household data, account access, stock records or billing are affected.
2. Contain
We take immediate steps to stop the harm: revoking sessions, rotating secrets, disabling an affected feature (for example, turning off AI import or billing webhooks), or taking the Service offline if necessary.
3. Investigate and fix
We use the audit trail, the inventory ledger and system logs to understand what happened and which households were affected, fix the root cause, and verify the fix.
4. Recover
If data must be restored, we restore from backups using our tested restore procedure, and reconcile any events (such as billing webhooks) received during the incident.
5. Notify
If an incident affects your personal information, we will notify affected household owners without undue delay and within the time required by applicable law, describing what happened, what data was involved, what we have done, and what you can do.
6. Learn
We write a post-incident review and update safeguards, tests and runbooks so the same failure doesn't recur.