Operating Agent Systems
Prompts, models and tool definitions are deployable production inputs. Versioning, evaluation before rollout, canaries judged on quality and cost, and a kill switch that does not need a redeploy.
Agent code, prompts, models and tool definitions all go through CI and an evaluation gate before they become an artifact, a deployment and then traffic.
Prompts, model identifiers, tool definitions and policies are versioned production inputs with a blast radius — not settings someone edits live.
Roll a behaviour change to a small share of traffic and judge it on quality, cost, latency and safety — where quality is the one that is not a threshold.
Disable a tool, disable the agent, or fall back to a simpler mode — all without a redeployment, because a redeploy is too slow when an agent is doing something harmful.
Per-request cost is a variable the system chooses at runtime, not a constant you can capacity-plan around — so budget ceilings become an operational control.