Cloud
- Reason explicitly about reliability, security, performance, cost, operations, and sustainability.
- Place stateful components and failure domains deliberately.
- Set recovery objectives and exercise restore paths before an incident.

Cloud · Original AlgoGym revision note with a source for deeper study.