Atlas / Skills / sangrokjung / Cc Dev Agent

Cc Dev AgentSAFE

skills/sangrokjung/cc-dev-agent

oh-my-zsh for Claude Code — 16 agents, 35 commands, 32 skills, 21 safety hooks in one install. v4.0 adds an adversarial review loop: a second agent that never sees the first one's reasoning. MIT.

Verdict
SAFE
Grade
B
Trust score
89 /100
Version
—
Hosts
1 documented
License
MIT
Stars
855
01

Overview

oh-my-zsh for Claude Code — 16 agents, 35 commands, 32 skills, 21 safety hooks in one install. v4.0 adds an adversarial review loop: a second agent that never sees the first one's reasoning. MIT.

Read from source at commit 31c3d29836edOBSERVED · 2026-10-12
02

Host compatibility

What the documentation claims. We have not run a compatibility test.

HostStatusNotes
claude-codementioned
03

What it tells the agent

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: cc-dev-agent
description: Use when starting Claude Code projects, writing CLAUDE.md/spec.md, dispatching subagents, or requesting Agent Teams parallel development. Covers Spec-Driven Development, Context Engineering, and post-dev workflow.
---

# Claude Code Agent

Claude Code를 활용한 Spec-Driven 개발 워크플로우 및 Context Engineering 가이드.

## 핵심 원칙

1. **Context Engineering**: 제한된 컨텍스트 윈도우에서 토큰 유용성 최적화
2. **Spec-Driven**: 코드 전에 스펙(spec.md) 먼저 → 반복 수정 8배 감소
3. **TDD Loop**: Red(테스트) → Green(구현) → Refactor
4. **Sub-agents 활용**: Explore, Plan 에이전트로 작업 분산
5. **작성-검증 분리**: `/clear` 후 새 눈으로 검증 (Boris 원칙)

---

## 병렬 개발 (Agent Teams v6)

> Git Worktree 기반 병렬 개발은 레거시. Agent Teams가 팀 생성/정리를 자동 처리.

### 전제조건

- `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS`를 `1`로 설정
- 설정 위치: `~/.claude/settings.json`의 `env` 또는 셸 환경변수

### 사용법

```bash
/orchestrate --type feature   # 기능 구현
/orchestrate --type bugfix    # 버그 수정
/orchestrate --type refactor  # 리팩토링
```

→ **상세 가이드**: `references/sub-agents.md`

---

## 개발 완료 후 워크플로우 (Boris 방식)

> "Give Claude a way to verify its work. If Claude has that feedback loop, it will 2-3x the quality."

### 핵심 플로우

```
/orchestrate (선택) → 개발 → /handoff-verify
→ /sync-docs (머지 전 문서 동기화 의무)
→ /commit-push-pr --merge
→ /web-checklist
```

### 단계별 역할

| 단계 | 명령어 | 이유 |
|------|--------|------|
| 1 | `/orchestrate` (선택) | Agent Teams 병렬 개발 |
| 2 | 개발 | 기능 구현 |
| 3 | `/handoff-verify` | 의도 문서화 + fresh context 자동 검증 (v6 통합) |
| 4 | `/sync-docs` | **머지 전 문서 동기화 (의무 게이트)** — 문서 변경분이 머지 커밋에 포함됨 |
| 5 | `/commit-push-pr --merge` | 커밋 & PR & 머지 (v7: sync-docs 게이트 내장 — 4단계 누락 시 자동 실행) |
| 6 | `/web-checklist` | 웹 테스트 체크리스트 |

→ **상세 가이드**: `references/post-dev-workflow.md`

---

## Context Engineering

| 명령어 | 용도 | 사용 시점 |
|--------|------|-----------|
| `/clear` | 컨텍스트 완전 초기화 | 새 Task, 에러 루프 탈출, **검증 전** |
| `/compact` | 대화 요약 압축 | 60% 도달 시 |
| `/context` | 현재 사용량 확인 | 복잡한 작업 중 수시 체크 |

→ **상세 가이드**: `references/context-engineering.md`

---

## Sub-agents

| 유형 | 용도 |
|------|------|
| `Explore` | 빠른 코드베이스 검색 (Read-only) |
| `Plan` | 구현 계획 설계 |

**병렬 처리**: 최대 10개 동시 실행

→ **상세 가이드**: `references/sub-agents.md`

---

## 개발 5단계

```
Phase 1: 초기화 → CLAUDE.md 생성
Phase 2: 기획 → spec.md, prompt_plan.md 작성
Phase 3: 구현 → TDD Loop 반복 (핵심)
Phase 4: 검증 → 스펙 대조, 코드 리뷰
Phase 5: 배포 → 보안 검토, 최종 점검
```

### Think 키워드

| 키워드 | 용도 |
|--------|------|
| `think` | 단순 버그 |
| `think hard` | 다중 파일, 원인 불명 |
| `ultrathink` | 아키텍처, 설계 결함 |

→ **상세 사용법**: `references/troubleshooting.md`

---

## Quick Reference

### 필수 .gitignore (병렬 개발 충돌 방지)

```gitignore
.claude/context/
.claude/settings.json
.claude/settings.local.json
.claude/handoff.md
CLAUDE.local.md
```

### 참조 문서

| 파일 | 용도 |
|------|------|
| `references/post-dev-workflow.md` | 개발 완료 후 워크플로우 상세 |
| `references/sub-agents.md` | 서브에이전트 + Agent Teams |
| `references/prompts.md` | 복사용 프롬프트 |
| `references/context-engineering.md` | 컨텍스트 엔지니어링 심화 |
| `references/playwright-agents.md` | Playwright 테스트 에이전트 |
| `references/templates.md` | MD 파일 템플릿 |
| `references/troubleshooting.md` | 에러 해결 + Think 키워드 상세 |
04

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codeNA
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
none-observed
Network
none-observed
Shell
none-observed
Dependencies
pinned
Secrets in source
none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.

Audited 2026-10-12 · audit v0.4.1 · source sha 31c3d29836edfull audit observations/trust-audit/skill/sangrokjung__cc-dev-agent.json · Report an issue / request a re-scan
05

Audit history

Every audit this skill has had.

DateSourceVerdictGradeScoreChange
2026-10-1231c3d29836edSAFEB89first audit
06

Questions

What does the Cc Dev Agent skill do?

oh-my-zsh for Claude Code — 16 agents, 35 commands, 32 skills, 21 safety hooks in one install. v4.0 adds an adversarial review loop: a second agent that never sees the first one's reasoning. MIT.

Is Cc Dev Agent safe to install?

The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.

What can Cc Dev Agent access on my machine?

The audit observed no filesystem, network or shell use at all in its source.

Which assistants does Cc Dev Agent work with?

Its documentation mentions claude-code. That is what the text claims, not a compatibility test we ran.

How current is this page?

The grade is for one exact copy of the source (31c3d29836ed), read on 2026-10-12. The repository is watched, and a new audit runs when it changes — this is the first audit.

Advertisement