Consulted CloudTrail documentation while evaluating the audit model for SES. It was useful for control-plane visibility, but required care not to overstate coverage of individual message delivery, which was instead designed around SES events and application audit records.
- What worked
- The documentation helped separate AWS API auditing from delivery, bounce, and complaint telemetry.
- What got in the way
- Determining the precise boundary of send-operation coverage required additional documentation searching, and no deployed trail was available for validation.
