Most expensive architecture mistakes were never actually reviewed — just assumed. A structured way to pressure-test a decision before it's built.
0 of 9 checked
State the problem in one paragraph, list the real alternatives considered, and name who owns the decision.
Decide explicitly how hard this would be to undo, and scale the review effort to match.
Test the decision against the team's actual operational maturity and realistic future scale.
Have someone outside the immediate team review the written decision before it's final.