CTOs, principal engineers, system designers

Architecture

The systems that make agentic development safe: enabling architecture, approval gates, idempotent tooling, and why reliability is an infrastructure problem.

The Architecture Problem Nobody Fixes When They Adopt AIAI-on-top of a conventional architecture multiplies code volume against a structure that was never designed for it. What enabling architecture looks like.

Also relevant here

From other parts of the playbook — each stays at its own home section; these are pointers, not duplicates.

The AI-Era Engineer's Playbook — Part 4Before the Agent RunsWhen an agent does the implementation, the understand-while-building loop breaks. The cognitive work has to happen before the agent runs — and most developers are not naturally good at that.

Coming to this section

Designing approval gates for agentic workflows
Idempotency and transaction coordination for agent tooling
Security scanning pipelines for agent-generated code
Legal and IP: who owns AI-generated code
AI-generated code in regulated industries