AI tools have crossed a line. They no longer just answer questions. They act.

Claude Code and Cowork are Anthropic's agentic AI tools: programs that can read your files, write code, execute tasks, send emails, and complete multi-step workflows on your behalf — without requiring you to write a single line of code. They are not chatbots. They are agents. That distinction matters enormously, and most users have not caught up to it yet.

-- Now includes coverage of Claude Cowork projects. --

Mastering Claude Code and Cowork is the practical guide to closing that gap. Elliott Middleton covers both tools from installation through advanced use: how the agentic loop works, how to direct Claude clearly, how to verify what it does, how to automate recurring workflows, and how to keep powerful tools in bounds when they are making decisions on your behalf. He does not oversell the technology. These tools hallucinate, make mistakes with files, and fail at tasks that should be easy. He covers that too — including the safety and permission practices that make sustained agentic use trustworthy rather than reckless.

The book is written for knowledge workers, not developers. Writers, analysts, project managers, lawyers, researchers, and executives will find concrete techniques they can use immediately. Developers will find chapters on Claude Code's deeper capabilities: MCP integrations, hooks, CI/CD pipelines, orchestrator-subagent architectures, and configuration hierarchies.

Twelve complete end-to-end workflows. Detailed coverage of Anthropic's subscription tiers. A clear-eyed assessment of where agentic AI is heading.

The gap between people who understand these tools and people who do not is widening. This book is designed to close it.