Agentic Coding with Claude Code: Build autonomous development workflows, orchestrate multi-agent systems MCP, and deploy AI-assisted pipelines that scale across engineering teams

Most developers use AI coding tools the same way they used autocomplete - one prompt at a time, one file at a time. The engineers shipping faster are doing something fundamentally different. They have turned Claude Code into a directed system with shared context, automated quality gates, and workflows the entire team benefits from. This is the book that shows you how to build that system. WHAT YOU WILL LEARN - Context engineering and CLAUDE.md - the shared infrastructure that makes every session smarter than the last- Directing vs prompting - task briefs, plan reviews, and session habits that produce reviewable code first time- Slash commands and hooks - reusable automated workflows your whole team runs consistently- Multi-agent orchestration - parallel agent systems that turn days of serial work into hours- Model Context Protocol - connect your AI workflows to GitHub, databases, Slack, and internal tools- Building custom MCP servers - full implementations in TypeScript and Python from scratch- Testing AI-generated code - the five failure modes and the quality loop that catches them before production- Team rollout and governance - API cost controls, security policy, and onboarding that works in days- Observability and deployment safety - session logging, audit trails, and production incident response WHO THIS BOOK IS FOR Senior software engineers, team leads, and engineering managers working in real codebases who want consistent, production-grade results. No prior experience with AI coding tools required. TypeScript and Python used throughout, all examples shown side by side.