Date: YYYY-MM-DD
- Review all postmortems from the last 30 days
- Check status of every remediation marked TODO or IN PROGRESS
- Update statuses (DONE / WONTFIX with reason / new deadline)
- Carry forward anything overdue with a realistic new deadline
- Any recurring tags across recent incidents? (e.g., three incidents in two months = systemic issue)
- Any recurring contributing factors? (e.g., "no alerting" showing repeatedly = need to prioritize monitoring)
- Any services with multiple incidents? (consider deeper investment)
- Were any runbooks used this month? Were they accurate?
- Update any runbooks that were wrong or incomplete
- Do any new common tasks need a runbook?
- One thing I can improve in the next hour based on this review (write it down and do it now)