Code review as teaching, not gatekeeping — note 86

The interesting constraint is not speed. It is whether code review as teaching, not gatekeeping — note 86 leaves the next person more capable.

Craft shows up in boring places: migrations sized to capacity, alerts that mean something, reviews that leave the code more teachable.

Incidents are expensive coaching. The write-up should change a checklist, a test, or an ownership map — not just a feeling.

Process should be light enough to change. If your AI workflow cannot be updated when a model, connector, or compliance rule changes, you do not have a workflow — you have a ritual.

Technical debt is not a moral failing. Unscheduled debt is. Put repayment on the same board as features.

Prefer reversible decisions. Architecture that cannot be walked back becomes politics.

In practice that means shorter cycles: decide, ship a thin slice, review what broke, coach the pattern into the next person. Long programs without those loops become status machines.

On code review as teaching, not gatekeeping — note 86, the leadership move is to make the invisible visible: ownership, verification, and the path for the next person.

When agents join the loop, treat them like junior systems: limited privileges, explicit tools, budgets, and a human who owns the outcome. Autonomy without audit is just distributed risk.

I prefer written decisions over verbal ones. Memory is a poor archive, and AI tools make fluent improvisation cheap — which raises the value of durable context.

None of this requires a new framework brand. It requires attention, a short feedback loop, and the humility to change process when agents join the workflow.

Ship the habit, not the slogan. Then measure whether the next person can run it without you.