Why AI-DLC?

AI-DLC is how software teams use the H•AI•K•U Method (Human AI Knowledge Unification). It provides a framework for organizing work into focused units with clear phases and responsibilities.

The Evolution of Software Development

The history of software engineering is a continuous quest to let developers focus on solving complex problems by abstracting away lower-level work. From machine code to high-level languages, from hand-rolled libraries to managed cloud services — each step moved humans further from implementation details and closer to problem expression.

Large Language Models marked a revolutionary shift, introducing conversational natural language for code generation, bug detection, and test creation. This was the AI-Assisted era — AI enhancing specific tasks while humans retained full control.

We have now entered the AI-Autonomous era. AI can maintain context across multi-hour workflows, iterate toward success criteria autonomously, and sustain the kind of reasoning that used to require entire teams. Iteration costs are approaching zero — you try something, it fails, you adjust, you try again, all in seconds rather than weeks.

Why a New Methodology

Traditional methods — Waterfall, Agile, Scrum — were designed for human-driven processes with long iteration cycles. Their sequential phases, handoff documentation, and approval gates made economic sense when changing requirements meant weeks of rework. But with AI, those phases are not just being augmented — they are collapsing into continuous flow.

Retrofitting AI into existing methods constrains its potential and reinforces outdated inefficiencies. AI-DLC is built from first principles for how development actually works now.

The Hat System

AI-DLC organizes work through hats — distinct mindsets that keep each phase of development focused. The default execution workflow uses three core hats, while specialized workflows add hats for security testing, design, TDD, and scientific debugging.

Planner

The Planner designs the implementation approach. This includes breaking work into steps, identifying dependencies, considering edge cases, and creating actionable plans. The output is a clear roadmap for the Builder to follow.

Builder

The Builder executes the plan. This is where code gets written, features get implemented, and tests get created. The Builder stays focused on the task at hand, following the plan without getting distracted by scope creep or tangential concerns.

Reviewer

The Reviewer validates quality and completeness. This includes checking that tests pass, requirements are met, edge cases are handled, and code quality meets standards. The Reviewer ensures work is ready for production.

Beyond the core three, AI-DLC includes specialized hats like Designer for UI/UX work, Red Team and Blue Team for adversarial security testing, and Observer, Hypothesizer, Experimenter, and Analyst for scientific debugging. See the full hat reference for details.

Units of Work

Work is organized into units. Each unit is a focused piece of functionality that can be completed in one session. Units have clear success criteria and acceptance tests. Breaking work into units ensures progress is measurable and momentum is maintained.

What This Enables

• Continuous flow: Phases collapse into a natural cycle rather than sequential gates
• Quality through backpressure: Automated enforcement guides AI toward quality, not manual review
• Human-on-the-loop: Define success criteria and guardrails, then let AI iterate toward convergence
• Near-zero iteration cost: Try, fail, adjust, and retry in seconds
• AI-native workflows: Built for how AI actually works, not retrofitted from human processes

Getting Started

AI-DLC is distributed as a Claude Code plugin. Install it in your project and start using the hat-based commands to structure your development workflow.

Part of the H•AI•K•U Method

AI-DLC is the software development profile of H•AI•K•U (Human AI Knowledge Unification) — a methodology for structured collaboration between humans and AI across any domain. H•AI•K•U provides the universal framework; AI-DLC applies it specifically to software development.

Go Deeper

For the full methodology — including production lessons, the Ralph Wiggum autonomous loop pattern, and the research behind AI-DLC — read the paper.

Part of Han

AI-DLC is part of the Han plugin ecosystem for Claude Code. Han provides a curated marketplace of plugins built on Bushido principles: quality, honor, and mastery.