[Book Review] "Claude Code Work Techniques for Business Professionals [Minolun Style]" - A Must-Read Guide for Engineers Who Use AI Every Day on How to Raise Your Own AI #minolun式

Hello, this is Usuda.

Everyone, are you raising your AI? (That's my greeting)

This time, I'm writing a book review of Minolun's book "［みのるん式］ビジネスパーソンのためのClaude Code仕事術" (Minolun-Style Claude Code Work Skills for Business Professionals).

https://gihyo.jp/book/2026/978-4-297-15939-9

As the title suggests, this book is designed so that even business professionals who aren't IT engineers can make use of it, but personally, I feel that engineers who use AI every day are precisely the ones who should read this book! (This bold text was added manually)

Why? Because if you're an IT engineer, I think you're probably working alongside many AI models and AI agents on a daily basis these days, and I suspect many of you feel that the switching costs between different tools and tasks are enormous. Whether you write code or not. I certainly feel that way myself.

In that context, about two months ago, when I had a conversation with Minolun, they taught me the "how to raise your own personal AI secretary" methodology that is compiled in this book. Since then, I've had a real sense that all sorts of things have been accelerating for me.

I'm sure that once you read this book, everyone will want to raise their own AI secretary! So let me introduce it.

By the way, while this book uses Claude Code for its explanations, the same knowledge can be applied to other AI agents and tools. I practice it with Kiro myself. After all, it's so cute!! kawaiiiiiiiiiiiiiiiiiiiiiiiiiiiiii!!!!!!!!!!!!!!!!!!

Book Overview

I'll introduce the highlights next, but let me start with an overview of the book.

• Book title: ［みのるん式］ビジネスパーソンのためのClaude Code仕事術
• Author: Onda Minoru ← Minolun
• Publisher: Gijutsu-Hyoronsha

Book Description (Quoted)

In recent years, the trend of leveraging AI in business—such as ChatGPT and Claude—has been rapidly advancing. And what is attracting attention right now is "AI agents," particularly "Claude Code" provided by Anthropic.

Unlike regular AI chat, AI agents can directly read and write files in your local PC environment and autonomously complete tasks. In other words, they can automate PC-based work such as complex research, document creation, and data processing. However, some of you may have the following questions.

I get that it's useful, but what exactly can it do?
Is it safe from a security standpoint?
What if I want to introduce it at my company but can't get approval?
In this book, the author—who uses Claude Code regularly and works as a technology consultant, presales professional, and training instructor in the cloud and AI fields—carefully explains the "work skills" cultivated on the front lines. By the time you finish reading, all of these questions will be resolved.

There are several Claude Code books aimed at business professionals, but this book's unique "3 strengths" are as follows:

Developing the habit of consulting AI about everything
Creating a "semi-automation" cycle
Leveraging the "compound interest of context" with a monorepo
By mastering these ［Minolun-style］ work skills, you can obtain your own personal AI secretary that "grows the more you use it." With your very own AI secretary, you can enjoy the greatest benefit of dramatically improving efficiency while maintaining the quality of your work.

"Not using it because I don't understand it" will lead to significant setbacks in the business environment going forward. Start by transforming the way you work with this one book, and gain a working style that is more human—one where you can spend time on the things you love!

Recommended for
Business professionals who want to incorporate Claude Code into their work
Especially those who want to promote AI adoption within their company but are stuck on security explanations and rule-making

Table of Contents

• Table of Contents
  • Introduction
• Chapter 1: Why Coding Agents Are Useful for Everyday Work
  • 1.1 Various Methods for Improving Work Efficiency
  • 1.2 Benefits of Using AI Agents to Improve Work Efficiency
  • 1.3 The Three Work Skills Proposed in This Book
• Chapter 2: Basic Knowledge of AI Agents
  • 2.1 What Exactly Is an AI Agent?
  • 2.2 Three Keywords That Are Key to This Book
  • 2.3 What Exactly Is Claude Code?
  • 2.4 Pricing Plans and the Concept of "Token Consumption"
  • 2.5 Basic Security Fundamentals to Understand First
• Chapter 3: Setting Up Claude Code and Basic Operations
  • 3.1 Before You Start Using Claude Code
  • 3.2 Setting Up Claude Code
  • 3.3 Getting a Feel for Basic Operations
  • 3.4 Checking Scheduling Candidates with "Connectors"
  • 3.5 Filling in Response Forms with "Browser Extensions"
• Chapter 4: Developing the Habit of Consulting AI About Everything
  • 4.1 Creating a Working Folder as the Foundation for Consultations
  • 4.2 Introducing Yourself by Voice and Passing Along Work Context
  • 4.3 Communicating Your Thoughts Without Organizing Them and Leaving the Organization to AI
  • 4.4 Consulting AI Before Starting Any Kind of Work
• Chapter 5 [Practice 1]: Let AI Handle the Pre-Work for Creating Materials
  • 5.1 Starting with a Consultation: "I Want to Create a Report"
  • 5.2 Gathering Scattered Materials and Compiling Them into a Single Document
  • 5.3 Brainstorming a Structure and Creating a Slide Framework File
  • 5.4 Finishing the Last 20% by Hand
• Chapter 6: Creating a "Semi-Automation" Cycle
  • 6.1 Semi-Automating with the "Human → AI → Human" Flow
  • 6.2 Adding the Necessary Tools
  • 6.3 Setting Up Guardrails to Prevent Mistakes
  • 6.4 Growing Procedure Files and Running an Improvement Cycle
• Chapter 7 [Practice 2]: Semi-Automating Expense Claims
  • 7.1 Consulting by Voice: "I Want to Do This Month's Expense Claims"
  • 7.2 Passing the Company's Expense Rules Along to AI
  • 7.3 Gathering and Cross-Referencing Scattered Application Materials
  • 7.4 Safely Processing Private Information from Card Statements
  • 7.5 Making It "Copy and Paste Only" Without Connecting to the System
• Chapter 8: Leveraging the "Compound Interest of Context" with a Monorepo
  • 8.1 Reasons for Consolidating Work into a Single Monorepo
  • 8.2 Organizing the Monorepo and Separating Confidential Projects
  • 8.3 Growing Your Own Personalized Settings
  • 8.4 Supplementing Functions with Tools and Sub-Agents
• Chapter 9 [Practice 3]: Accelerating Project Work with "Compound Interest"
  • 9.1 Creating the Project's "Current Status" from Scattered Materials
  • 9.2 Updating the Current Status and Tasks Based on Meeting Content
  • 9.3 Turning Research Assignments Born from Meetings into Decision-Making Material
  • 9.4 Creating Everything from Research Results to a Report Slide
  • 9.5 Turning Successes and Stumbles into Skills and Rules
• Chapter 10: Expanding Your Developed Workspace in Four Directions
  • 10.1 Improving Output Quality with the Concept of "Specification-Driven Development"
  • 10.2 Eliminating the "AI-Generated Feel" Common in AI Text
  • 10.3 Safely Working in Parallel Across Multiple Sessions
  • 10.4 Issuing Instructions from a Smartphone Even When Away from the Office
  • 10.5 Turning Your Developed Workspace into a Team Asset
  • 10.6 Using Your Developed Workspace with Other AI Agents
• Chapter 11: How to Overcome the "Security Wall" in Organizations
  • 11.1 Creating a Path for Safe Use Rather Than Prohibition
  • 11.2 Advancing Implementation in Cooperation with the Security Department
  • 11.3 Determining the Data to Protect and the Scope of Operations
  • 11.4 Starting Internal Adoption with Low-Risk Semi-Automation
  • 11.5 Preparing for Traps Planted from Outside and Human Complacency
  • 11.6 Organizing the Usage Environment for Organizations with Corporate Contracts and Management Settings
  • 11.7 Growing Guardrails That Balance Utilization and Safety

It's a fairly dense 296 pages in A5 size.

Highlights

Throughout each chapter, you can feel the practical knowledge that Minolun has accumulated, but here I'll introduce just three points I particularly want to recommend.

Throw Everything at AI and Consolidate It in One Place

"Chapter 4: Developing the Habit of Consulting AI About Everything" explains how to consolidate tasks that are being thrown to various AI tools in various places, making everything work within a single location.

Normally, the AI tools and places where AI agents are run tend to be divided by project or work boundaries, working directories, and file locations, but practical techniques are introduced for how to consolidate all of these into one place to make things easier to use and work with.

Things like "it's OK to input things roughly by voice" and "let AI handle all the folder organization" — each individual technique is introduced along with Minolun's personal experiences, presenting concrete techniques, practical methods, and ways of thinking.

Looking at the items alone, you might wonder if the bar seems a little high or whether things will go well, but I think reading Minolun's writing will motivate you with a "I want to try this myself! Let's go for it!" feeling.

Semi-Automation Techniques for Running AI at Full Speed Without Fear

"Chapter 6: Creating a 'Semi-Automation' Cycle" compiles practical methods for leveraging automation to use AI without fear.

Recently, I think people are increasingly having AI agents work autonomously, but on the other hand, many people may not be able to fully leave things to AI because having everything executed automatically without oversight is problematic.

Techniques for semi-automation that connect to running things at full speed are introduced, such as how to add tools to integrate and how to set up guardrails.

In particular, the roles of each guardrail in Claude and methods for using AI to adjust the guardrails themselves can be used to eliminate the fear of delegating work to AI, and the criteria for adjusting guardrails based on underlying principles that aren't limited to Claude will also be useful.

Consolidating Context to Create Compound Interest

"Chapter 8: Leveraging the 'Compound Interest of Context' with a Monorepo" shows a more advanced way to utilize the information consolidated in Chapter 4, laying out the path toward a true AI secretary in action that breaks free from fragmented AI usage.

The switching costs that arise from switching between AI tools and tasks can be handled seamlessly by always consulting an AI agent in the same location, allowing you to continue or resume work seamlessly — it truly becomes the state of "this is the AI I wanted!"

This is the part I've personally been experiencing over the past two months, and I want everyone to feel the importance of consolidating into one place.

However, as you consolidate more and more information, there are more things to be mindful of, so this chapter also covers how to organize your harness and how to handle confidential information. Moving forward while maintaining this balance is important, and this chapter will be a major hint for doing so.

Summary

I've introduced Minolun's book "［みのるん式］ビジネスパーソンのためのClaude Code仕事術."

Since I have the personal experience of practicing what I learned from conversations with Minolun, I understand the effectiveness of the information compiled in this book.

Once again, I really want engineers who use AI every day to pick this up.

https://gihyo.jp/book/2026/978-4-297-15939-9