Claude Code Production Engineering: Build Agents and Modernize Legacy Systems

Modernize the system without guessing what it does

Legacy modernization rarely fails because a team cannot generate enough new code. It fails because nobody can prove which behavior matters, who still depends on it, or whether a replacement is safe to operate.

Claude Code for Legacy Modernization is a practical field handbook for changing critical systems without pretending that the repository is a complete specification. It shows you how to recover hidden business rules, reconstruct runtime behavior, identify unknown consumers, and turn those findings into evidence that engineers and decision owners can review.

The method follows one bounded modernization slice from discovery to retirement.

You will learn how to:

recover business rules from code, data, runtime flows, side effects, and operational knowledge; establish a trustworthy baseline before asking AI to change anything;
build characterization tests, reviewed golden masters, contract tests, invariants, and differential comparisons;
find or introduce a safe implementation seam;
create a reversible strangler path;
compare legacy and replacement behavior before cutover;
define acceptance criteria and rollback triggers in advance;
cut over under named human authority;
measure accepted outcomes and retire the old path without losing the evidence trail.

Claude Code assists with repository discovery, test preparation, bounded implementation, old-and-new comparison, contradiction checks, and evidence assembly. It does not decide which behavior is correct or authorize a production change. Architects, domain owners, security reviewers, and operational leaders keep those decisions.

The book follows Northstar Settlements, a fictional financial-services system, through a complete modernization thread. The companion lab lets you reproduce the baseline, inspect the legacy behavior, run the tests, expose an intentional replacement mismatch, and complete the exercises in your own branch.

What is included

• more than 790 pages of practical guidance;
• 26 connected modernization chapters;
• 86 architecture, workflow, and evidence diagrams;
• a runnable GitHub exercise repository;
• synthetic fixtures, a retained golden master, and differential tests;
• Claude Code project controls and compatibility wrappers;
• reusable schemas, checklists, evidence records, and decision templates.

Companion lab

Northstar Modernization Lab

This book is for software engineers, architects, technical leads, quality engineers, and platform or operations professionals responsible for systems that are valuable, difficult to understand, and dangerous to change.

### Continue with the companion guides

Need the wider production-agent engineering foundation? Claude Code: Building Production Agents That Actually Scale covers tools, MCP, hooks, permissions, evals, observability, cost engineering, and multi-agent operations.

Need both books? The Claude Code Production Engineering bundle combines that production-agent guide with this complete legacy-modernization handbook.

A Best-Selling Book on LeanPub for weeks since releasing. A mandatory guide for building Claude production AI agents.

This is a Claude Code production agents book for engineers who need agents that can run with tools, permissions, MCP servers, evals, observability, and cost controls in real systems.

Most Claude Code tutorials stop at "hello world." This book covers what happens after that: when your agent needs to run reliably in production, at scale, in environments where failure has real consequences.

Written by an AI engineer who builds Claude Code agent systems for regulated financial institutions, it walks through the full production stack.

Part I covers the agent loop, context management, and model selection.

Part II builds the primitive layer: tools, hooks, skills, MCP servers, and plugins.

Part III bridges the CLI to the Claude Agent SDK, covering the dispatch loop, session management, and tool registration for headless deployment. Part IV tackles governance: permissions, sandboxing, secrets, audit trails, and managed settings.

Part V covers evals, LLM-as-judge patterns, observability, cost engineering, and failure modes.

Part VI puts it together with team workflows, deployment patterns, multi-agent orchestration, and a full walkthrough of Anthropic's open-source financial services reference agents (the production-grade agent templates Anthropic released in May 2026).

Thirty-one chapters, five practical appendices (including a ninety-day production-readiness checklist, an eval starter kit, and an MCP server audit template), and code extracted from production systems.

If you are a senior AI engineer, technical lead, or architect evaluating Claude Code for production use, this is the reference that will save you months of trial and error.

Continue with the companion guides

Modernizing a difficult existing system? Claude Code for Legacy Modernization turns the production disciplines in this book into an evidence-led method for discovery, characterization testing, reversible migration, and controlled cutover.

Need both books? The Claude Code Production Engineering bundle combines this production-agent guide with the complete legacy-modernization handbook.

Responsible for security approval, architecture governance, or production risk? Securing Enterprise AI Agents covers bounded autonomy, AgentSecOps, MCP security, secure RAG, evaluation gates, and the controls required when agents can act.