AI-generated code fails in a predictable set of ways, and predictable means catchable. Get the working version of the register and score your last AI-assisted release against it this week.

​

The review checklist. All 16 failure modes, grouped by the four causes behind them, with the question to ask and the controls that catch each one.

The one-page scoring sheet. Score the control that actually ran on your last AI-assisted change, from 0 (not considered) to 4 (gated). Critical gaps stand out at a glance.

The scoring workbook. The same sheet in Excel, with the gap flags and averages by cause worked out for you, plus every source behind every row.