skills.rest
kenjudy
kenjudyCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

pdca-framework

Enforces PDCA governance and strict TDD for AI-assisted coding sessions.

41|5|Updated Aug 22, 2025
Search Tags:#code quality#TDD#software development#developer workflow#human-AI collaboration#AI coding#PDCA
One-click install
npx skills add https://github.com/kenjudy/pdca-code-generation-process --skill pdca-framework
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pdca-framework
Source: https://github.com/kenjudy/pdca-code-generation-process/tree/main/claude-skill/src
Command: npx skills add https://github.com/kenjudy/pdca-code-generation-process --skill pdca-framework

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

AI code generation, while powerful, often leads to increased technical debt, duplicated code, and decreased delivery stability. This Skill provides a human-supervised PDCA (Plan-Do-Check-Act) framework that addresses these challenges by enforcing structured development, strict Test-Driven Development (TDD), and continuous human oversight. It ensures that AI-generated code maintains high quality, reduces rework, and keeps developers actively engaged and accountable.

Core Features & Use Cases

  • Structured PDCA Workflow: Guides AI through systematic Plan, Do, Check, and Act phases for every coding task, ensuring thorough analysis, disciplined implementation, rigorous validation, and continuous improvement.
  • Strict TDD Enforcement: Integrates mandatory Test-Driven Development, requiring the AI to write failing tests first, then minimal code to pass them, significantly reducing bugs and improving code quality.
  • Active Human Oversight & Intervention: Empowers developers to act as "process police," intervening immediately when the AI deviates from established working agreements or TDD discipline, maintaining control and accountability.
  • Use Case: A software team needs to integrate a complex new feature. They use this Skill to guide their AI assistant through each PDCA phase, from initial architectural analysis and detailed planning to test-driven implementation and final validation. This ensures the AI produces high-quality, well-tested code that aligns with existing patterns and minimizes technical debt.

Quick Start

To begin a new development task, instruct Claude to analyze the problem using the PDCA framework. For example: Analyze implementing a new user authentication module using the PDCA framework. Search the codebase for existing patterns before proposing an approach.

Frequently Asked Questions about pdca-framework

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reduce technical debt when using AI for code generation?▼

Technical debt from AI coding stems from lack of oversight and discipline. The PDCA framework enforces structured planning, strict Test-Driven Development, and human validation at each phase, ensuring AI-generated code maintains quality, respects existing architecture, and minimizes rework and duplicated code.

What's the best way to implement Test-Driven Development with AI coding assistants?▼

TDD with AI requires the assistant to write failing tests first, then minimal code to pass them. The PDCA framework makes this mandatory across Plan, Do, Check, and Act phases, giving developers active oversight to intervene when the AI deviates from TDD discipline and maintain code quality.

How can I maintain code quality and developer accountability in AI-assisted development?▼

Active human oversight is critical in AI coding workflows. The PDCA framework positions developers as process police, empowering them to intervene immediately when the AI breaks established working agreements or architectural patterns, ensuring accountability and preventing quality degradation.

Can I use PDCA governance for iterative AI coding across production codebases?▼

Yes. The PDCA framework is designed specifically for iterative AI coding sessions in production environments. It guides each nested cycle through planning, implementation, validation, and retrospection phases while enforcing TDD, respecting existing architecture, and requiring explicit human oversight.

What should I do before starting an AI-assisted coding task?▼

Start by analyzing the problem using the PDCA framework's Plan phase. Search the codebase for existing patterns and architectural conventions, then propose an approach aligned with current design. This preparation ensures the AI understands constraints and context before implementation begins.

Why does human-supervised AI coding reduce delivery instability?▼

Unsupervised AI generates code that often duplicates existing patterns and increases technical debt. Human-supervised PDCA enforces systematic validation, continuous improvement cycles, and TDD discipline, ensuring generated code integrates cleanly with existing systems and maintains delivery reliability.

Related Skills

Explore More Software Engineering Skills→
markeriksonmarkerikson
41

code-style-review

Document coding conventions from configuration files and source code examples.

Community
Intermediate
erhankaraarslanerhankaraarslan

monitoring

Automate production health checks, metrics collection, and alerting with Prometheus and Grafana.

Community
Basic
luongldptitluongldptit

springboot-verification

Verify Spring Boot projects for merge and deployment readiness.

Community
Advanced
ogghstogghst

pm

Update sprint and product backlogs and track iteration progress in Backcast EVS.

Community
Basic