fithia2/dev/phase7_deliverables/operator_runbook.md

Phase 7 Operator Runbook

Purpose

This runbook describes the human operating procedure for controlled live rollout.

Daily Pre-Market Checklist

• Confirm current live stage and approved capital limits.
• Review unresolved incidents.
• Review overnight drift alerts.
• Confirm active release/config ids match approved versions.
• Confirm all mandatory sources met freshness SLA.
• Confirm no-trade or degraded mode status.

During Market Hours

• Monitor new alerts.
• Review exceptional approval tickets.
• Do not change stage or release unless emergency procedure is invoked.
• If severe incident occurs, prioritize containment over diagnosis.

End-of-Day Checklist

• Confirm reconciliation completed.
• Review fill quality and slippage outliers.
• Check post-trade reviews created for all closed trades.
• Review freeze or rollback recommendations.
• Verify daily report generated and archived.

Emergency No-Trade Procedure

1. Activate no-trade mode.
2. Verify new orders are blocked.
3. Decide whether open non-protective orders should be canceled.
4. Record incident id and operator note.
5. Notify approver/governance group.

Stage Freeze Procedure

1. Confirm freeze condition evidence.
2. Record stage freeze event.
3. Prevent further promotions.
4. Continue or reduce trading according to risk policy.
5. Schedule review before lifting freeze.

Stage Rollback Procedure

1. Confirm rollback trigger.
2. Move to prior approved stage.
3. Update active capital and risk caps.
4. Record change and approver.
5. Review all open positions for compliance.

Post-Incident Procedure

• Open incident record if not already created.
• Collect affected trade ids.
• Link config/release versions.
• Assign root-cause owner.
• Schedule corrective and preventive actions.