Monozukuri — AI skill for Claude Code
An AI coding skill that applies Japanese-inspired craftsmanship to build simple, correct, verifiable, and maintainable software.
How to install Monozukuri
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open kuya-egg/Monozukuri and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Monozukuri does
An AI coding skill that applies Japanese-inspired craftsmanship to build simple, correct, verifiable, and maintainable software.
Alternatives in AI
- AI Dev Tasks — A simple task management system for managing AI dev agents 7.8k ★
- Agent Rules Books — AGENTS.md rules / skills for AI coding agents: Codex, Cursor & Claude Code 2.6k ★
- Riteway — Simple, readable, helpful unit tests 1.2k ★
README
monozukuri
`monozukuri` is a complete family of Agent Skills for building and maintaining software as a craft rather than disposable output.

It combines Japanese-inspired quality principles with practical software engineering: YAGNI, KISS, clear design, defect prevention, testing, security, observability, documentation, and evidence-based review.
It is designed for both greenfield construction and existing codebases. The target is maintainable software—not theoretically perfect software: code that future maintainers can understand, verify, operate, and change safely.
The skill is deliberately **Japanese-inspired, not Japanese-prescriptive**. It does not claim that developers in one country share a single engineering style. The useful ideas are treated as metaphors for disciplined software work.
The craftsmanship system
The umbrella skill applies this loop:
Observe → Understand → Plan → Craft → Verify → Reflect → Kaizen
The repository also contains focused, independently installable skills:
| Skill | Focus |
|---|---|
nemawashi |
Clarify intent and prepare the design |
genchi-genbutsu |
Inspect the real repository and environment |
kanso |
Choose simple, durable designs |
kata |
Implement small, integrated vertical slices |
jidoka |
Debug failures and find root causes |
poka-yoke |
Prevent mistakes and design meaningful tests |
kodawari |
Review details that carry quality |
andon |
Make health and failure visible |
kaizen |
Improve code incrementally |
hansei |
Learn from failures and completed work |
shukka |
Release, migrate, and hand off responsibly |
What it changes
When active, an agent should:
- clarify the problem before conse
Related Skills
Services New Software Skill
A Claude Code skill that applies Sequoia's "Services: The New Software" framework to evaluate AI product ideas
Tausik Core
AI coding agents can't quietly fake "done". Signed ed25519 receipts prove the gates actually ran — bound to th
Willow
Willow — a companion narrator that gives blind developers situational awareness while an AI coding agent (e.g.
Magent Multi AI Agents Manager
Launch each project's AI coding agent (Claude Code, Codex, Cursor) in its own terminal and auto-tile every win
Single File Wbs
単一HTMLで動く WBS/ガント+イナズマ線ビューア。データはJSON1枚・Claude Codeで保守 / Single-file WBS & Gantt viewer with inazuma (progress)
FastAPI Agent Blueprint
Build FastAPI AI agents with a clear blueprint for reliable, maintainable app architecture
Related Agents
Swe Dev High Bar Engineering
PyTorch-style high-bar engineering. Produces maintainable changes that improve the system; adapts behavior acr
Ark Prototyper
Use this agent to build minimal, verifiable prototypes of Ark features. It makes the smallest changes needed t
Lw Scribe
Mechanical, diff-verifiable edits. Use for renames, formatting, boilerplate, import fixes and repetitive chang