Agentic Reliability Engineering

Chapter 12. Simulation-Driven Reliability

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be the 12th chapter of the final book. Please note that the GitHub repo will be made active later on.

If you’d like to be actively involved in reviewing and commenting on this draft, please reach out to the editor at [e-mail].

Simulation is the cheapest place to discover what an agent will do. A bad decision in production costs minutes of incident, hours of investigation, days of postmortem, and weeks of recalibration. A bad decision in simulation costs the compute to run it. The orders-of-magnitude difference in cost is what makes simulation the single most leveraged investment in the agentic reliability stack, and the chapter argues that this leverage is not optional. Every promotion to a higher authority tier passes through a simulator first. Every novel action class is exercised against a scenario library before earning production authority. Every change to the architecture’s reasoning substrate is validated against simulated outcomes before being trusted in production. The simulator is not a development convenience. It is the substrate that makes safe autonomy possible.

The framing worth installing upfront is that simulation is not testing. Testing asks “does this code ...