Agentic Reliability Engineering

Chapter 11. Governed Autonomy

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 11th 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].

The most operationally consequential chapter of this book argues for a single claim. Governance is the operating system of autonomy. Without it, autonomy is just unsupervised execution. With it, autonomy is something the team can grant, measure, narrow, and revoke as evidence accumulates. The architecture’s Governance Plane from Chapter 4.6 is the substrate this chapter develops into a runtime control system. The autonomy boundary from Chapter 1.5 is the engineered surface the system honours by design. What this chapter installs is the operational discipline that connects the architectural primitives to the daily work of an SRE running an agentic system.

This chapter is the most “Monday-morning” chapter in the book, by which we mean: most of what an SRE actually changes about a production agentic system lives here. The four planes from Chapter 4 are stable architectural commitments the team made once. The five flagship SLOs are tracked continuously but adjusted ...