Operations

These runbooks describe the implemented protected workflow definitions and the operator procedures around them. The definitions are locally validated but disabled/unconfigured; no real provider has been operated.

• Protected deployment
• Drift management
• Rollback
• Destroy and expiry
• Incident response
• Troubleshooting

Operating principles

• Treat every change as an auditable proposal with an exact commit and plan hash.
• Prefer read-only diagnosis and reversible action.
• Stop at provider, state, identity, budget, or approval mismatch.
• Keep AWS and Azure runbooks distinct after the shared decision points.
• Record what was observed, who approved it, what changed, and how health was verified.