1. Review the situation.
Collect a safe summary of the symptoms, timing and recent changes. Name the person responsible for the system and the limits of the investigation.
Process
Start with one important failure.
Define the task, the people affected and the response you need.
2. Follow the signals.
Review existing checks and diagnostic information. Identify what distinguishes a healthy result from a misleading success or a missed failure.
3. Agree the response.
Choose a useful alert, recipient and first action. Document escalation and recovery decisions, including when a specialist is needed.
4. Test carefully.
Use an isolated scenario or agreed exercise to check detection and notification. Do not trigger customer-facing failures or contact third parties without authorization.
5. Hand over the instructions.
Record what was tested, what remains uncertain and who maintains the checks. A controlled test does not prove every future incident will be caught.