Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
npx claudepluginhub tumblecat44/20loc-jackpot --plugin codeloopInteractive project setup wizard for Codeloop — the AI autonomous development loop engine. Fixed stack: Vercel + Railway + Next.js + FastAPI + PostgreSQL. Focuses on product planning (what to build, target customer, revenue model, MVP scope), then validates ALL required env vars are ready before allowing `codeloop start`. Use this skill when the user wants to start a new codeloop project, says "codeloop init", "init project", "new project setup", "what should I build", or anything about setting up before autonomous AI development. Also trigger when someone asks about project templates for AI-driven coding, or mentions wanting to build a SaaS/web app with AI.
AI 자율 개발 루프 제어 스킬. codeloop start/stop/status 서브커맨드로 Stop Hook 기반 무한 루프를 시작·중지·모니터링한다. Triggers: codeloop, codeloop start, codeloop stop, codeloop status, 루프 시작, 루프 중지, 루프 상태, ループ開始, ループ停止, 循环启动
Fetch accurate, up-to-date API documentation via context-hub (chub) CLI instead of relying on stale training data. Automatically activated when working with any external library, framework, or API. Use this skill BEFORE writing any code that uses an external library or API. This ensures you use current, correct APIs — not hallucinated ones.
codeloop start 한 번이면 목표 LOC까지 AI가 알아서 개발 · sleep · 재개를 반복한다.
nowimslepe 프로젝트에서 311K LOC 달성으로 검증된 Ralph Loop 엔진을 범용 Claude Code Plugin으로 추출.
"codeloop.yaml 20줄 + PROMPT.md 한 장 = 20만 줄의 프로덕션 코드"
# 1. Marketplace 등록
/plugin marketplace add https://github.com/tumblecat44/20loc-jackpot.git
# 2. 플러그인 설치
/plugin install codeloop@tumblecat44-20loc-jackpot
또는 대화형 UI에서:
/plugin
# → Discover 탭 → codeloop 선택 → 설치 범위(user/project/local) 선택
설치하면 Stop Hook이 자동 등록된다. settings.local.json을 직접 편집할 필요 없음.
# 1. 프로젝트 셋업 (3문답 위저드)
/codeloop-init
# 2. 루프 시작
/codeloop start
# 3. 상태 확인
/codeloop status
# 4. 루프 중지
/codeloop stop
codeloop start
→ claude -p (PROMPT.md 주입)
→ Claude 자율 개발
→ Stop Hook 자동 발동
├─ Usage Gate (OAuth API → sleep or pass)
├─ Completion Gates (loc/iterations/budget/custom)
├─ Dashboard 갱신 (loc-status.md)
└─ 게이트 통과? → 루프 종료 : exit 0 → 반복
.claude/codeloop.state.md 존재 = 루프 ON, 삭제 = OFFfiveHourPercent 실측값으로 정확한 rate limit 관리project:
name: "My SaaS"
prompt: ./PROMPT.md
model: opus
gates:
- type: loc
target: 200000
- type: iterations
max: 500
usage:
threshold: 90
cooldown_seconds: 30
sleep_buffer_seconds: 60
dashboard:
path: .claude/loc-status.md
| Gate | Example | Description |
|---|---|---|
loc | target: 200000 | 누적 LOC 목표 도달 시 종료 |
iterations | max: 500 | 반복 횟수 제한 |
budget | max_usd: 50 | API 비용 한도 |
custom | command: "./my-gate.sh" | 사용자 스크립트 |
codeloop/
├── .claude-plugin/
│ ├── plugin.json # Plugin meta + Stop Hook 자동 등록
│ └── marketplace.json # Marketplace 등록 정보
├── hooks/
│ └── stop-hook.sh # Orchestrator
├── scripts/
│ ├── usage-gate.js # OAuth API 사용량 감지
│ ├── dashboard.sh # Markdown 대시보드
│ ├── parse-yaml.sh # YAML 파서
│ └── gates/
│ ├── loc.sh # LOC gate
│ ├── iterations.sh # Iterations gate
│ ├── budget.sh # Budget gate
│ └── custom.sh # Custom gate
└── skills/
├── codeloop/
│ └── SKILL.md # /codeloop start|stop|status
└── codeloop-init/
├── SKILL.md # /codeloop-init 위저드
└── references/
├── stacks.md
└── prompt-template.md
MIT
Share bugs, ideas, or general feedback.
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
loophaus — Control plane for coding agents
Autonomous session orchestrator for Claude Code - manages multi-phase development workflows
Universal CLI orchestrator with multi-runner support. Autonomous spec-driven development with dependency DAG, parallel worktree execution, two-stage review gates, and modular merge hardening.
連續自引用 AI 循環,用於互動式迭代開發。實現 Ralph Wiggum 技術,讓 Claude 在 while-true 循環中執行相同提示直到任務完成
Autonomous multi-agent development framework with spec-driven sprints and convergent iteration
Auto-loop execution workflow with quality gates for Claude Code. Automatically decomposes tasks, implements code, runs tests, and iterates through quality gates until completion.
Blocks Korean input and reviews your English at the end of every response. Learn English by using Claude Code!
AI가 유저에게 기생하지 않도록 자율성 강박을 심는 플러그인. 라이트 형제는 자전거 도구로 비행기를 만들었다. 아폴로는 4KB로 달에 갔다. 너에게는 컴퓨터에서 사람이 할 수 있는 99%의 도구가 있다. 3-Strike Rule: 최소 3가지 시도 후에만 질문 허용.
터미널 AI Agent들의 끝장 토론 플러그인
Automatic context engineering - project conventions accumulate and refine as you code
4-agent build team for Claude Code. Tell it what to build — PM writes the PRD, engineer implements, QA tests. Boundaries enforced by hooks, wave-based execution, research-backed design from 18 sources.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claim