• 0
• 0
• 0

Summary

This audiobook is narrated by an AI Voice. 
Most writing about AI agents is either hype or abstract philosophy. This book is neither. It is an opinionated, practical guide to building production agents with Claude Code, written by an engineer who has spent the last decade building automation systems for financial institutions, where a failure can mean a regulator audit. 
 
It is for senior engineers, technical leads and architects who need to ship agents into production and keep them running. 
 
The book is organised in six parts. 
 
Part I, Orientation: the agent loop, what Claude Code is, the model family, context as a first-class resource and a working definition of production. 
 
Part II, The Primitive Stack: tools, subagents, hooks, slash commands, skills, MCP, plugins and a compliance monitoring agent built end to end. 
 
Part III, From CLI to SDK: when to leave the CLI, the two SDK layers, custom orchestrators, sessions, state and durability. 
 
Part IV, Governance and Security: the permissions model, sandboxing, network egress, secrets, policy as code and the supply chain. 
 
Part V, Evals, Observability, Reliability: the eval mindset, evals for agentic workloads, LLM-as-judge, observability and failure modes. 
 
Part VI, Production Patterns and the Road Ahead: cost engineering, team workflows, anti-patterns and legacy migration, a measured look ahead and Anthropic's financial services reference agents. 
 
This narration follows the 8 October 2026 edition, which reviews Opus 5.5, Fable 5.1, Sonnet 5.5 and Haiku 5.5 against official documentation from that date. Prices and limits are dated references, so check the current documentation before you deploy. 
 
The appendices are not narrated.

Duration: about 16 hours (16:20:40)
Publishing date: 2026-10-30; Unabridged; Copyright Year: — Copyright Statment: —

• Science & Technology