The Agent Baseline defines 35 controls across six security outcomes—but the right starting point depends on how your organization uses agents. Learn how to sequence controls for coding, internal, and production agents.