메뉴
Domain-Driven Development workflow specialist using ANALYZE-PRESERVE-IMPROVE cycle for behavior-preserving code transformation. Use when refactoring legacy code or reducing technical debt safely.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
AI text humanization and 윤문 (post-editing) specialist that detects and removes AI tells while preserving meaning, facts, and figures. Covers Korean, English, Japanese, and Chinese with a shared severity model (S1/S2/S3), quality grades (A/B/C/D), and 30%/50% over-editing guardrails. Use to make AI-generated text read as human-authored without changing what it says (de-ai, naturalness pass).
MoAI unified orchestrator for autonomous development. Routes natural language or subcommands (plan, run, sync, project, fix, loop, mx, feedback, review, clean, codemaps, gate, harness) to specialized agents.
Markdown-to-single-file-HTML report renderer. Six modes (status, incident, plan, explainer, financial, pr) selected automatically by report type. Zero external JS/CSS framework dependencies — inline SVG charts, single font-CDN exception for Korean readability. Self-contained output for email attachment, print, and offline viewing.
Unified thinking toolkit: Creative frameworks (Critical Evaluation, Diverge-Converge, Deep Questioning), First Principles reasoning (absorbed from moai-foundation-philosopher), and Adaptive Thinking via the `ultrathink` keyword. Use for ideation, strategic analysis, architecture decisions, and deep reasoning on complex problems.
Provides MoAI-ADK foundational principles including TRUST 5 quality framework, SPEC-First DDD methodology, delegation patterns, progressive disclosure, agent catalog reference, and token budget management (absorbed from moai-foundation-context). Use when referencing TRUST 5 gates, SPEC workflow, or context window optimization.
Provides MoAI-ADK foundational principles including TRUST 5 quality framework, SPEC-First DDD methodology, delegation patterns, progressive disclosure, agent catalog reference, and token budget management (absorbed from moai-foundation-context). Use when referencing TRUST 5 gates, SPEC workflow, or context window optimization.
| name | moai-workflow-ddd |
| description | Domain-Driven Development workflow specialist using ANALYZE-PRESERVE-IMPROVE cycle for behavior-preserving code transformation. Use when refactoring legacy code or reducing technical debt safely. |
| license | Apache-2.0 |
| compatibility | Designed for Claude Code |
| allowed-tools | Read, Write, Edit, Bash(git:*), Bash(pytest:*), Bash(ruff:*), Bash(npm:*), Bash(npx:*), Bash(node:*), Bash(uv:*), Bash(make:*), Bash(cargo:*), Bash(go:*), Bash(mix:*), Bash(bundle:*), Grep, Glob, mcp__context7__resolve-library-id, mcp__context7__get-library-docs |
| user-invocable | false |
| metadata | {"version":"1.0.0","category":"workflow","status":"active","updated":"2026-01-16","modularized":"true","tags":"workflow, refactoring, ddd, domain-driven, behavior-preservation, ast-grep, characterization-tests","author":"MoAI-ADK Team","related-skills":"moai-tool-ast-grep, moai-workflow-testing, moai-foundation-quality"} |
[NOTE] This workflow is selected based on .moai/config/sections/quality.yaml:
constitution:
development_mode: ddd # or tdd
When to use this workflow:
development_mode: ddd → Use DDD (this workflow)development_mode: tdd → Use TDD instead (moai-workflow-tdd)Key distinction:
Domain-Driven Development provides a systematic approach for refactoring existing codebases where behavior preservation is paramount. Unlike TDD which creates new functionality, DDD improves structure without changing behavior.
Core Cycle - ANALYZE-PRESERVE-IMPROVE:
When to Use DDD:
When NOT to Use DDD:
Greenfield Project Adaptation:
For new projects without existing code, DDD adapts its cycle:
This makes DDD a superset of TDD - it includes TDD's test-first approach while also supporting refactoring scenarios.
TDD Approach (for new features):
DDD Approach (for refactoring):
The golden rule of DDD is that observable behavior must remain identical before and after refactoring. This means:
The analyze phase focuses on understanding the current codebase structure and identifying refactoring opportunities.
Identify logical boundaries in the codebase by examining:
Use AST-grep to analyze structural patterns. For Python, search for import patterns to understand module dependencies. For class hierarchies, analyze inheritance relationships and method distributions.
Evaluate code quality metrics:
Low cohesion and high coupling indicate refactoring candidates.
Use AST-grep to identify problematic patterns:
Create analysis reports documenting:
The preserve phase establishes safety nets before making any changes.
Characterization tests capture existing behavior without assumptions about correctness. The goal is to document what the code actually does, not what it should do.
Steps for creating characterization tests:
Characterization test naming convention: testcharacterize[component]_[scenario]
For complex outputs, use snapshot testing to capture current behavior:
Snapshot files serve as behavior contracts during refactoring.
Before proceeding to improvement phase, verify:
Run mutation testing if available to verify test effectiveness.
The improve phase makes structural changes while continuously validating behavior preservation.
Never make large changes at once. Follow this pattern:
Extract Method: When a code block can be named and isolated. Use AST-grep to identify candidates by searching for repeated code blocks or long methods.
Extract Class: When a class has multiple responsibilities. Move related methods and fields to a new class while maintaining the original API through delegation.
Move Method: When a method uses data from another class more than its own. Relocate while preserving all call sites.
Inline Refactoring: When indirection adds complexity without benefit. Replace delegation with direct implementation.
Rename Refactoring: When names do not reflect current understanding. Update all references atomically using AST-grep rewrite.
Use AST-grep for safe, semantic-aware transformations:
For method extraction, create a rule that identifies the code pattern and rewrites to the extracted form.
For API migration, create a rule that matches old API calls and rewrites to new API format.
For deprecation handling, create rules that identify deprecated patterns and suggest modern alternatives.
After each transformation:
When executing DDD through moai:2-run in DDD mode:
Step 1 - Initial Assessment:
Step 2 - Analyze Phase Execution:
Step 3 - Preserve Phase Execution:
Step 4 - Improve Phase Execution:
Step 5 - Validation and Completion:
For complex refactoring requiring multiple iterations:
Behavior Preservation (Required):
Structure Improvement (Goals):
Apply TRUST 5 framework with DDD focus:
DDD relies heavily on AST-grep for:
Load moai-tool-ast-grep for detailed pattern syntax and rule creation.
DDD complements testing workflow:
DDD outputs feed into quality assessment:
Tests Fail After Transformation:
Characterization Tests Are Flaky:
Performance Degradation:
When DDD session encounters issues:
Version: 1.0.0 Status: Active Last Updated: 2026-01-16
| Rationalization | Reality |
|---|---|
| "This legacy code is obviously broken, I will just rewrite it" | The broken-looking code may encode production behavior callers depend on. ANALYZE first, rewrite with tests. |
| "Characterization tests are just busywork" | They are the safety net that lets you refactor without fear. Without them, refactor equals rewrite. |
| "I understand this module well enough to skip ANALYZE" | Understanding is not verification. ANALYZE forces you to name assumptions and confront callers. |
| "This function has no tests because it has no behavior worth testing" | If it has callers, it has contract. The contract is the thing to characterize. |
| "I will preserve behavior by being careful" | Carefulness is not a mechanism. Tests are the only mechanism that catches regressions. |
| "This refactor is too small to need DDD" | Small refactors cause silent regressions exactly because people skip the safety net. |
Chesterton's Fence: Do not remove or replace code whose purpose you cannot explain. ANALYZE phase exists specifically to uncover the reasoning behind code that looks wrong.