recover-from-a-work-mistake

# recover-from-a-work-mistake Protect affected people and systems first. Accountability means accurate status, appropriate escalation, repair, and durable prevention. ## Procedure 1. Stop unsafe or compounding activity within the person's authority and preserve relevant evidence. 2. Determine the observed impact, affected scope, time window, current state, and what remains unknown. 3. Notify the correct owner or incident channel promptly with facts, actions taken, and needed decisions. 4. Follow the applicable safety, security, privacy, financial, legal, or operational response process. 5. Separate the error, contributing conditions, downstream effects, and individual responsibility. 6. Build and execute an authorized repair plan with verification and rollback where relevant. 7. Communicate corrections to affected people without excuses, speculation, or promises outside authority. 8. Identify why controls failed and choose preventive changes at the responsible layer. 9. Test the repair and prevention, document residual risk, and close only with owner acceptance. ## Failure plan - Never delete logs, alter records, conceal impact, blame others speculatively, or make an unauthorized admission of liability. - Do not delay disclosure until the full cause is known. - If health, safety, security, privacy, money, or legal obligations may be involved, escalate immediately. - Avoid punishment-focused analysis that ignores system conditions and control gaps. ## Done - Containment, scope, notification, and repair are recorded with timestamps - Affected outputs or records are reconciled and the fix is verified - Preventive action has an owner, test, due date, and accepted residual risk