Practical Agentic Coding: Beyond Vibe Coding and AI Assistants for Successful Autonomous Software Development
暫譯: 實戰 Agentic Coding:超越 Vibe Coding 與 AI 助理,實現成功的自主軟體開發

Patil, Abhijeet, Patil, Manisha

  • 出版商: Apress
  • 出版日期: 2026-11-08
  • 售價: $1,930
  • 貴賓價: 9.5 折 $1,833
  • 語言: 英文
  • 頁數: 697
  • 裝訂: Quality Paper - also called trade paper
  • ISBN: 9798868826696
  • ISBN-13: 9798868826696
  • 相關分類: AI Coding、Large language model
  • 海外代購書籍(需單獨結帳)

相關主題

商品描述

Software development teams worldwide are struggling with a paradox: AI tools promise increased productivity, but many implementations result in technical debt, security vulnerabilities, and workflow disruptions. The problem isn't AI capability, but the lack of proper orchestration and governance. Practical Agentic Coding solves this challenge by introducing a strategic, structured approach to autonomous AI development.

This book teaches leaders and experienced developers on how to transition from using AI as a passive assistant to leading agents as autonomous team members. You will journey through a complete evolution: from setting a foundational industry vision and evaluating the market value of tools like Amazon Kiro, Google Antigravity, and Claude Code, to architecting intelligent workflows where AI agents handle complex tasks while humans provide strategic direction.

By combining the best architectural practices with hands-on tutorials, security frameworks, and real-world case studies, this practical guide enables teams to harness AI's transformative power without compromising quality or control. The result: development teams that operate at unprecedented speed and scale while maintaining the craftsmanship that defines exceptional software.

You Will Learn To:

  • Understand the core principles of Agentic Coding and how to use AI Agents not just as tools but valuable team members
  • Design workflows where AI agents can handle tasks like coding, testing, documentation, and more with human oversight.
  • Evaluate the unique value propositions of leading tools like Kiro, Antigravity, Cursor, GitHub Copilot, and Claude Code
  • Implement guardrails, audit trails, and governance practices to ensure safe and reliable agent behavior.
  • Learn how to balance autonomy and control within integrated human/AI teams.
  • Recognize and prevent common pitfalls like bad code generation, security issues, and broken workflows caused by poorly managed agents.

This Book is for:

  • Developers and Engineers looking to gain the skills required to transition from writing code to orchestrating high-performance agentic teams.
  • Tech Leaders who want to gain a strategic insight to evaluate the market, manage AI-driven risk, and design an autonomous delivery organization.

商品描述(中文翻譯)

全球各地的軟體開發團隊正面臨一個矛盾:AI 工具承諾能提升生產力,但許多導入方案卻導致技術債、安全漏洞與工作流程中斷。問題不在於 AI 的能力,而在於缺乏適當的協調與治理機制。《Practical Agentic Coding》透過引入策略性且結構化的方法,協助讀者解決自主式 AI 開發所面臨的挑戰。

本書教導技術領導者與資深開發人員,如何從將 AI 當作被動助理,轉變為帶領 AI agents 擔任自主團隊成員。您將完整走過這段演進歷程:從建立產業基礎願景,以及評估 Amazon Kiro、Google Antigravity 和 Claude Code 等工具的市場價值,到設計智慧型工作流程,讓 AI agents 負責處理複雜任務,而人類則提供策略方向。

本實務指南結合最佳架構實務、動手操作教學、安全性框架與真實世界案例研究,協助團隊在不犧牲品質或控制力的前提下,充分發揮 AI 的變革力量。最終,開發團隊將能以前所未有的速度與規模運作,同時維持造就優秀軟體的精湛工藝。

您將學會:

• 了解 Agentic Coding(代理式程式設計)的核心原則,以及如何將 AI Agents 不只是當作工具,更視為有價值的團隊成員
• 設計工作流程,讓 AI agents 在人類監督下處理程式撰寫、測試、文件撰寫等任務
• 評估 Kiro、Antigravity、Cursor、GitHub Copilot 和 Claude Code 等領先工具的獨特價值主張
• 實作防護機制(guardrails)、稽核軌跡(audit trails)與治理實務,確保 agents 的行為安全且可靠
• 學習如何在整合人類與 AI 的團隊中,取得自主性與控制力之間的平衡
• 辨識並預防常見問題,例如錯誤的程式碼生成、安全性問題,以及因 agents 管理不當所造成的工作流程失效

本書適合:

• 希望培養必要技能,從撰寫程式碼轉型為協調高效能 agentic 團隊的開發人員與工程師
• 希望建立策略性洞察力,以評估市場、管理 AI 驅動的風險,並設計自主交付組織的技術領導者

作者簡介

Abhijeet Ganpatrao Patil is a Senior Technology Leader, IT Architect, and Software Engineer with over two decades of experience driving digital innovation on a global scale. In his previous role as an Executive Architect and Hybrid Cloud Leader at IBM, he served as a principal advisor to IT leaders and global enterprise customers, orchestrating multi-year IT transformations and complex cloud adoption programs. Currently a Senior Solutions Architect at AWS based in Raleigh, NC, Abhijeet leverages his background as a software engineer and his Executive MBA in Cybersecurity to help customers navigate their Cloud and AI journeys. He is a core member of the AI and Machine Learning Technical Field Community at AWS and an active contributor to AWS Labs' open source agentic coding frameworks.

Manisha Patil is an MBA graduate specializing in Cloud Management and Data Analytics with a Bachelor's in Instrumentation and Controls Engineering. Specializing in the integration of cloud and AI tooling from providers such as GitHub and OpenAI, she focuses on building intelligent, scalable applications that prioritize human-centered design. With experience in Industrial Automation and an MBA in Cloud Management, Manisha brings a unique blend of technical precision and strategic insight to the development of agentic workflows, ensuring that AI-driven systems enhance rather than replace human craftsmanship.

作者簡介(中文翻譯)

Abhijeet Ganpatrao Patil 是一位資深技術領導者、IT 架構師與軟體工程師,擁有超過二十年的經驗,致力於在全球推動數位創新。在擔任 IBM Executive Architect 與 Hybrid Cloud Leader 期間,他為 IT 領導者及全球企業客戶提供主要諮詢,負責規劃並推動多年期 IT 轉型,以及複雜的雲端採用計畫。目前 Abhijeet 任職於 AWS,駐於北卡羅來納州 Raleigh,擔任 Senior Solutions Architect。他憑藉軟體工程師背景,以及 Cybersecurity Executive MBA 學位,協助客戶規劃並推進 Cloud 與 AI 之旅。他是 AWS AI and Machine Learning Technical Field Community 的核心成員,也是 AWS Labs 開源代理式(agentic)程式設計框架的積極貢獻者。

Manisha Patil 擁有 MBA 學位,專攻 Cloud Management 與 Data Analytics,並具備 Instrumentation and Controls Engineering 學士學位。她專精於整合 GitHub、OpenAI 等供應商提供的雲端與 AI 工具,致力於建構以人為本、智慧化且可擴充的應用程式。Manisha 具備 Industrial Automation 經驗,並擁有 Cloud Management MBA 學位,為代理式(agentic)工作流程的開發帶來技術精準度與策略洞察兼具的獨特能力,確保 AI 驅動的系統能夠提升而非取代人類的專業創作。

最後瀏覽商品 (20)