A reflection on why agent safety is moving from prompt-level guardrails to governed runtimes like NVIDIA OpenShell.

Key architecture patterns from Stripe's Minions autonomous coding agents for building reliable AI agents in production.

Using traces to gain visibility into agent behavior across development, testing, and production.

An overview of state-of-the-art AI-driven software engineering practices.

A reflection on what could be a precisely calibrated and controllable organization of contexts provided to different agents assigned code-writing tasks.