Press release

Open-Source Approach to Ensuring Quality with AI — Classmethod Releases AI-Assisted Test-Driven Development Framework “Tsumiki”

Classmethod, Inc. (Headquarters: Minato-ku, Tokyo; CEO: Satoshi Yokota;”Classmethod”) has officially released “Tsumiki,” an open-source framework designed to systematize and streamline the use of Anthropic’s AI agent-driven coding tool, “Claude Code.”

■ Background and Challenges

In recent years, the rise of AI-powered “vibe coding”—a development method where code is generated based on natural language instructions—has made it easier for even those with limited programming experience to build systems. However, in corporate development environments, the following serious issues have become apparent:

  • Not implemented as expected — Frequent rework due to ambiguous specifications
  • Works, but seems hard to maintain — Accumulation of technical debt due to lack of testing
  • Fine for simple demos, but risky for production — Lack of quality assurance processes

Over the past 12 months, Classmethod has worked with more than 40 client companies on AI-assisted development and discovered that solving these issues requires not only “pursuing smarter AI” but also “discipline in the development process.”

Additionally, through the use of Claude Code, many companies have expressed the following needs:

  • Adopt a systematic approach to Test-Driven Development (TDD)
  • Build a standardized workflow across the team
  • Establish a development process that ensures quality while leveraging AI
  • Eliminate disparities in individual prompt engineering skills

To meet these needs, Classmethod has systematized its accumulated expertise in AI-assisted test-driven development and released it as a framework.

■ Overview and Features of Tsumiki

Tsumiki enables the construction of reliable production-grade systems through a new development method called AI-assisted Test-Driven Development (AITDD), based on the principles of Specification First, Test First, and Quality First.

It goes beyond simply generating “working code” and enables efficient development of “clearly specified, test-protected, and maintainable production-quality systems” by fully leveraging AI capabilities.

Positioning of Tsumiki

Within the AI-driven development ecosystem, Tsumiki is positioned as follows:

Claude (AI model) → Claude Code (middleware) → Tsumiki (framework) → Industry Templates → Individual Applications

Tsumiki serves as a bridge to effectively utilize the powerful features of Claude Code in real-world development.

Differences Between AITDD and Conventional AI-Driven Development

AITDD combines traditional Test-Driven Development (TDD) with AI technology to enhance development efficiency.

Basic Development Cycle:

Requirements Expansion → Design → Task Breakdown → TDD Implementation
(Requirements Analysis → Write Test Case → Red → Green → Refactor → Verify)

Features of the Extended TDD Cycle:

Red (Write Test): Define the “expected behavior” as a test first
Green (Implementation): AI generates the minimal code needed to pass the test
Refactor (Improvement): AI optimizes the code for readability and efficiency
Verify (Validation): AI confirms that the implementation meets the requirements

Differences from Conventional AI-Driven Development:

Tsumiki ensures quality not just by generating code with AI, but also through:

  • Automatic Specification Generation and Management: Clarifies requirements and ensures consistency with implementation
  • Automated Code Review: Objective quality evaluation by AI

■ Key Components of Tsumiki

Tsumiki consists of the following main components:

1．Kairo Command Set: Provides a comprehensive development flow from requirements expansion to implementation

  • Automates requirements expansion, design, task breakdown, and implementation
  • Ensures consistent process management

2．TDD Command Set: Supports each step of test-driven development with AI

  • Implements the Red-Green-Refactor-Verify cycle
  • Automatically evaluates based on quality standards

3．Integrated Workflow

  • Automates development processes by combining multiple commands
  • Dedicated AITDD command sets (/kairo-* series, /tdd-* series)

■ Public Information

  • License: MIT License (Commercial use permitted)

■ Future Outlook

Tsumiki is also planned to be expanded with industry-specific templates. Specialized templates tailored to the characteristics of various industries—such as finance, healthcare, manufacturing, and retail—will be provided sequentially.

About Classmethod, Inc.

Classmethod, Inc. is a technical partner that supports corporate digital transformation (DX) with a focus on cloud-native technology areas, including Amazon Web Services (AWS), data analysis, mobile, IoT, and AI/machine learning. In terms of AWS support, Classmethod has been recognized as a top-tier partner continuously since 2015 and has won the “AWS Consulting Partner of the Year” (awarded to the top partner in Japan) five times. In 2022, it received the global accolade of “SI Partner of the Year,” and was named a finalist again the following year in 2023, establishing its position as a world-class AWS partner in both name and substance.

To date, Classmethod has supported approximately 5,600 companies and managed/built a cumulative total of over 40,000 AWS accounts. The company also heavily emphasizes a culture of technical knowledge sharing among its engineers, having published over 60,000 technical articles on its owned media “DevelopersIO.” Additionally, it operates “Zenn,” a knowledge-sharing platform for engineers, contributing further to the development of the tech community. Guided by the philosophy of “continuously contributing to the creative activities of all people,” Classmethod proposes optimal technologies that enhance its customers’ business value.

0120-991-668 Contacts