Agentic AI with Claude Code : A practical guide to design, build, and ship production-grade AI agents across Anthropic ecosystem

  • 予約

Agentic AI with Claude Code : A practical guide to design, build, and ship production-grade AI agents across Anthropic ecosystem

  • ただいまウェブストアではご注文を受け付けておりません。 ⇒古書を探す
  • 製本 Paperback:紙装版/ペーパーバック版
  • 言語 ENG
  • 商品コード 9781808655357

Full Description

Build production-grade Claude agents with a clear decision framework spanning the API, Agent SDK, Claude Code, MCP, Skills, and no-code paths.

Key Features

Choose the right Anthropic build path using a practical decision framework
Build and operate agents across the API, Agent SDK, Claude Code, MCP, Skills, and no-code paths
Apply a four-layer model that stays useful as the Anthropic ecosystem evolves
Purchase of the print or Kindle book includes a free PDF eBook

Book DescriptionChoose the right way to build an AI agent before committing time and code. This practical guide gives you a four-layer model of the Anthropic ecosystem and a repeatable framework for choosing between the Claude API, Agent SDK, Claude Code, MCP, Skills, and no-code paths.
You'll start with the foundations shared by every agent, including models, loops, tools, memory, workflows, and evaluations. From there, you'll build agent loops with the API, implement workflow patterns and RAG, create production agents with the Agent SDK, and expose reusable capabilities through MCP servers. You'll then compose agents in Claude Code using CLAUDE.md, Skills, commands, subagents, hooks, orchestration patterns, schedules, CI, and plugins. A dedicated no-code path shows you how to work with Projects, Cowork, connectors, and team rollouts. Throughout, three builders- a product engineer, a platform developer, and a non-coding founder- carry the same project down each path, showing concepts from 3 angles.
Written by Chorouk Malmoum, Founder and CTO of AgentX Academy and a LinkedIn Top Voice in AI, the book combines hands-on agent building with practical guidance drawn from teaching thousands of professionals.
By the end, you'll be able to design, build, deploy, and operate Claude agents while confidently choosing the approach that fits each problem.What you will learn

Understand models, loops, tools, and memory in AI agents
Choose between API, Agent SDK, Claude Code, and no-code paths
Build agent loops and reusable workflow patterns with Claude
Ship production agents with the Claude Agent SDK
Build MCP servers for reusable Claude capabilities
Compose Claude Code agents with Skills, subagents, and hooks
Run agents through schedules, CI, plugins, and team workflows
Create no-code agents with Projects, Cowork, and connectors

Who this book is forDevelopers, product engineers, tech leads, founders, operators, and technically curious professionals who want to build and ship Claude agents using code, configuration, or no-code approaches. Basic AI tool familiarity is sufficient; code-focused sections assume basic TypeScript or Python and command-line skills.

Contents

Table of Contents

The Map: How Anthropic Wants You to Build Agents
The Engine Room: Models, Money, and the Agent Loop
What You Get for Free: Built-in Tools, Evals, MCP and Skills
Choose Your Path
Building the Loop by Hand
Workflows and Knowledge: Patterns, RAG, and Agentic Search
The Agent SDK: Never Write the Loop Again
Building MCP Servers: Ship Capabilities, Not Agents
Production: Evals, Cost, Guardrails - and the Headless Shortcut
Claude Code Is an Agent Platform: The Runtime and Its Memory
Skills and Commands: Teaching the Agent Your Way
Subagents and Hooks: Workers and Guarantees
Orchestration: Guardrails, Reach, and Multi-Agent Systems
Unattended and Shipped: Schedules, CI, and Plugins
Agents Without Code: Projects and Cowork
The No-Code Agent Platform: Skills, Schedules, and Connectors
Surfaces, Teams, and Graduation

最近チェックした商品