「코더」 — 개발 방법론 담당
“코드를 짜 달라"는 요청과 “소프트웨어를 만들어 달라"는 요청은 다릅니다. 전자는 한 번의 답변으로 끝나지만, 후자에는 요구사항 정리, 테스트, 품질 검사, 문서화라는 공정이 필요합니다. 코더는 이 공정 전체를 가진 직원입니다. 집을 지을 때 목수 한 명이 아니라 설계도(SPEC)·시공(run)·준공 검사(sync)를 갖춘 시공팀을 부르는 것과 같습니다.
핵심은 MoAI-ADK 개발 방법론의 무설치 실행입니다. moai CLI를 따로 설치하지 않아도 Claude Code나 Claude Desktop에서 /moai 명령으로 SPEC 기반 개발(plan — 요구사항 명세, run — DDD/TDD 구현, sync — 문서 동기화)을 그대로 쓸 수 있습니다. TDD(테스트를 먼저 쓰고 통과시키는 개발 방식)와 DDD, 품질 게이트, 백엔드·프론트엔드·데이터베이스·보안 참조 지식까지 29종 스킬이 들어 있어, 비개발자도 자연어로 개발 워크플로우를 시작할 수 있습니다.
이 페이지는 요약입니다. 코더의 실제 사용법은 Code 가이드에서, 심화 개발 문서(SPEC 워크플로우·에이전트 구조·품질 프레임워크 전체)는 adk.mo.ai.kr/ko에서 다룹니다.
flowchart LR
A["요청
(/moai plan)"] --> B["SPEC 작성
요구사항 명세"]
B --> C["run
DDD/TDD 구현"]
C --> D["품질 게이트
테스트·린트 검사"]
D --> E["sync
문서 동기화"]
E --> F["산출물
(동작 코드 + 문서)"]방법론(moai-workflow-*), 도메인 지식(moai-domain-*), 참조 패턴(moai-ref-*) 계열 29종의 전체 목록입니다.
스킬 29종 — 아래 표는 마켓플레이스 정본(v6.2.0)에서 자동 생성됩니다.
| 스킬 | 무엇을 해주나 |
|---|---|
claude-agentic-coding | 개발자용 Claude agentic 코딩 — tool use, sub-agents, MCP 연동으로 자율 코딩 워크플로우 구성. Claude 공식 tool-use/sub-agents 문서 기반. "Claude로 자율 코딩 에이전트 만들고 싶어" 질문에 즉시 활용. |
moai-domain-backend | Backend development specialist covering API design, database integration, microservices architecture, and modern backend patterns. Use when designing APIs, implementing server logi |
moai-domain-database | Database specialist covering PostgreSQL, MongoDB, Redis, Oracle, and cloud database platforms (Neon, Supabase, Firestore). Use for schema design, query optimization, indexing strat |
moai-domain-frontend | Frontend development specialist covering React 19, Next.js 16, Vue 3.5, and modern UI/UX patterns with component architecture. Use when building web UIs, implementing components, o |
moai-domain-html-report | 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 dep |
moai-domain-humanize | 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 |
moai-foundation-cc | Canonical Claude Code authoring kit covering Skills, sub-agents, plugins, slash commands, hooks, memory, settings, sandboxing, headless mode, and advanced agent patterns. Use when |
moai-foundation-core | Provides MoAI-ADK foundational principles including TRUST 5 quality framework, SPEC-First DDD methodology, delegation patterns, progressive disclosure, agent catalog reference, and |
moai-foundation-quality | Code quality orchestrator enforcing TRUST 5 validation, proactive code analysis, linting standards, and automated best practices. Use for code review, quality gate checks, or TRUST |
moai-foundation-thinking | Unified thinking toolkit: Creative frameworks (Critical Evaluation, Diverge-Converge, Deep Questioning), First Principles reasoning (absorbed from moai-foundation-philosopher), and |
moai-harness-learner | Harness learning subsystem coordinator. Produces Tier 4 auto-update proposal payloads consumed by the orchestrator (which surfaces them via AskUserQuestion) and orchestrates Apply/ |
moai-meta-harness | DEPRECATED — legacy 7-Phase meta-harness. Redirects to the v4 harness Builder (/moai:harness <natural-language request>) which replaces the static 7-Phase workflow with an orchestr |
moai-ref-api-patterns | REST/GraphQL API design patterns, error handling conventions, and input validation reference for backend development. Agent-extending skill that amplifies backend domain work (spaw |
moai-ref-git-workflow | Git workflow patterns, branch strategies, conventional commits, and PR templates reference for git operations. Agent-extending skill that amplifies manager-git expertise with produ |
moai-ref-llm-security | AI/LLM defensive security reference: prompt-injection defense, OWASP LLM Top 10 defensive mapping, MCP and agentic tool-call hardening, training-data poisoning detection, model-out |
moai-ref-owasp-checklist | OWASP Top 10 security checklist, authentication patterns, input validation, and HTTP security headers reference. Agent-extending skill that amplifies backend-implementation and sec |
moai-ref-react-patterns | React/Next.js component design patterns, state management strategies, and project structure reference for frontend development. Agent-extending skill that amplifies frontend domain |
moai-ref-secops | DevSecOps, container, and API operational defensive security reference: CI/CD pipeline hardening, secret scanning, IaC misconfiguration detection, SAST/DAST integration, container |
moai-ref-supply-chain | Software supply-chain defensive security reference: SBOM generation and verification (SPDX / CycloneDX), dependency-confusion defense, malicious-package triage playbook, SLSA prove |
moai-ref-testing-pyramid | Test pyramid strategy, coverage targets, test patterns, and quality metrics reference. Agent-extending skill that amplifies manager-develop test-creation and quality-validation wor |
moai-workflow-ci-loop | Unified CI watch + auto-fix loop skill. Polls gh pr checks after /moai sync PR creation, classifies required vs auxiliary failures, attempts safe automated patches (max 3 iteration |
moai-workflow-ddd | Domain-Driven Development workflow specialist using ANALYZE-PRESERVE-IMPROVE cycle for behavior-preserving code transformation. Use when refactoring legacy code or reducing technic |
moai-workflow-loop | Ralph Engine - Automated feedback loop with LSP diagnostics and AST-grep integration for continuous code quality improvement. Use when implementing error-driven development, automa |
moai-workflow-project | Integrated project management system covering documentation, language initialization, template optimization, docs generation, and JIT document loading. Absorbed from moai-workflow- |
moai-workflow-spec | SPEC workflow orchestration with EARS format requirements, acceptance criteria, and Plan-Run-Sync integration for MoAI-ADK development. Use when creating SPEC documents or defining |
moai-workflow-tdd | Test-Driven Development workflow specialist using RED-GREEN-REFACTOR cycle for test-first software development. Use when developing new features from scratch or when behavior speci |
moai-workflow-testing | Comprehensive testing and development workflow specialist combining DDD testing, characterization tests, performance profiling, and quality assurance. Use when writing tests or mea |
moai-workflow-worktree | Git worktree management for parallel SPEC development with isolated workspaces, automatic branch registration, and seamless MoAI-ADK integration. Use when setting up parallel devel |
moai | MoAI unified orchestrator for autonomous development. Routes natural language or subcommands (plan, run, sync, project, fix, loop, mx, feedback, review, clean, codemaps, gate, harn |
코더에는 code-investigator(읽기 전용 조사 에이전트)가 포함되어 코드베이스의 의존성·공개 API·영향 범위를 조사해 구조화 보고를 올립니다. 구현하는 손과 조사하는 눈을 분리해, 수정이 미치는 범위를 먼저 파악한 뒤 코드를 고치게 하는 구조입니다.
| 에이전트 | 역할 | 유형 |
|---|---|---|
code-investigator | 코드베이스 조사 서브에이전트 — 의존성·공개 API·영향 범위를 조사해 구조화 보고. claude-agentic-coding 워크플로우에서 위임. 사용자 프롬프트 불가 (blocker report 반환). | 검수 (읽기 전용) |
1. 기능 하나를 SPEC부터. “회원 가입 기능 만들어줘"라고 하면 /moai plan이 요구사항을 SPEC 문서로 정리해 확인받고, /moai run이 TDD로 구현하며, /moai sync가 문서를 맞춰 줍니다. 무엇이 만들어질지 코드 작성 전에 합의된다는 점이 핵심입니다.
2. 레거시 코드 안전 리팩토링. “이 오래된 코드 정리하고 싶은데 망가질까 봐 무서워"라고 요청하면 DDD 워크플로우(moai-workflow-ddd)가 현재 동작을 테스트로 고정한 뒤 동작을 보존하며 구조만 개선합니다.
3. 배포 전 품질 점검. “커밋 전에 품질 검사 돌려줘"라고 하면 품질 게이트가 린트·포맷·타입 검사·테스트를 병렬로 실행해 통과 여부를 보고합니다.
잘 안 될 때 — /moai 명령이 인식되지 않으면 플러그인 설치 상태부터 확인하세요(플러그인 가이드). 워크플로우 자체가 낯설다면 Code 가이드의 입문 문서를 먼저 읽고, 심화 내용은 adk.mo.ai.kr/ko를 참조하시면 됩니다.
- context7 — 라이브러리·프레임워크의 최신 공식 문서를 조회하는 문서 검색 통로. 학습 시점 이후 바뀐 API도 최신 문서 기준으로 구현하게 해 줍니다. 별도 자격증명 없이 동작합니다.