AI coding assistant toolkit - git workflow protection, session handoff, context management, and development automation
npx claudepluginhub baleen37/bstack --plugin bstackAutomate browser interactions, test web pages and work with Playwright tests.
Use when creating, organizing, or improving Claude Code rules files in .claude/rules/ — modular project or user-level instructions that Claude automatically loads
Use when the user asks to "commit", "git commit", "커밋", or runs /commit — reviews changes, stages files, and creates a single git commit
Use when designing systems, architectures, or APIs, when multiple valid approaches exist with no single obvious answer, or when user explicitly requests parallel/competing solutions
Create PR — commit, push, PR, wait for merge.
Use when changes span multiple components, services, or layers and unit/integration tests alone cannot confirm the full flow works correctly
Use when comparing two variants (code, LLM prompts, CLI commands, or any executable) against defined criteria with the same inputs. Do NOT use when variants cannot produce observable, comparable output.
Use when a PR is broken and needs to be fixed — CI failures, merge conflicts, BEHIND base branch, or failing tests.
Use when setting up a new project repository, initializing CI/CD, configuring branch protection, or the user asks to scaffold project infrastructure
You are a QA engineer AND a bug-fix engineer. Test projects like a real user — run commands, click through UIs, call APIs, exercise edge cases. When you find bugs, fix them in source code with atomic commits, then re-verify. Produce a structured report with before/after evidence.
This skill should be used when the user asks "how should I...", "what's the best approach...", "do you remember...", "why did we...", mentions "last time", "before", "we discussed", or when stuck after investigation. Searches conversation history using claude-mem 3-layer workflow.
Use when exploring unfamiliar codebases, investigating bugs, or learning new technologies before acting
Review recent conversations to find improvements for CLAUDE.md files.
Use when you need to run multiple AI CLI agents (claude, codex, gemini) in parallel on independent subtasks — spawns visible tmux split panes for real-time monitoring
AI 코딩 어시스턴트 툴킷 — Claude Code, OpenCode, 그 외.
bstack은 플러그인 형태로 묶인 단일 통합 패키지입니다:
--no-verify 등 위험한 git 명령어 자동 차단# Add this repository as a marketplace
claude plugin marketplace add https://github.com/baleen37/bstack
# Install the plugin
claude plugin install bstack
bstack/
├── plugins/
│ ├── me/ # Core personal workflow plugin
│ │ ├── hooks/ # Session hooks (git guard, handoff, LSP checks)
│ │ └── skills/ # Personal skills (16 total)
│ ├── ralph/ # Ralph Loop plugin
│ │ ├── hooks/ # Ralph persistence hooks
│ │ └── skills/ # ralph, ralph-cancel
│ ├── autoresearch/ # Autonomous experiment loop plugin
│ │ ├── commands/
│ │ ├── hooks/
│ │ └── skills/ # autoresearch
│ ├── jira/ # Jira integration plugin
│ │ └── skills/ # 5 Jira workflow skills
│ ├── core/ # Shared agent definitions
│ ├── lsp-*/ # Individual LSP plugins (bash, go, lua, etc.)
├── scripts/ # Utility scripts (handoff, dispatch, version sync)
├── docs/ # Development and testing documentation
├── tests/ # BATS tests
├── schemas/ # JSON schemas
└── CLAUDE.md # Project instructions for Claude Code
me plugin (personal workflow)| Skill | Description |
|---|---|
handoff | 세션 종료 시 다음 세션을 위한 인계 문서 생성 |
pickup | 이전 세션의 인계 문서 로드 |
create-pr | 커밋, 푸시, PR 생성 통합 워크플로우 |
pr-pass | CI 실패, 머지 충돌 등 깨진 PR 수정 |
commit | Conventional Commits 형식으로 커밋 |
research | 코드베이스 탐색 및 버그 조사 |
e2e | 다수 컴포넌트에 걸친 E2E 검증 |
eval-harness | 두 변형을 정의된 기준으로 비교 평가 |
iterate | 반복 단일 변경 사이클로 점진적 개선 |
competitive-agents | 병렬 경쟁 에이전트로 설계 탐색 |
tmux-workers | tmux 분할 창에서 병렬 AI 에이전트 실행 |
claude-code-rules | .claude/rules/ 규칙 파일 생성/관리 |
remembering-conversations | 이전 대화 컨텍스트 검색 및 적용 |
review-claudemd | CLAUDE.md 개선사항 발굴 |
reddit-fetch | WebFetch 차단 시 Reddit 콘텐츠 가져오기 |
ralph plugin| Skill | Description |
|---|---|
ralph | PRD 기반 자동 반복 개발 루프 실행 |
ralph-cancel | 실행 중인 Ralph 루프 취소 |
autoresearch plugin| Skill | Description |
|---|---|
autoresearch | 최적화 목표를 위한 자율 실험 루프 실행 |
jira plugin| Skill | Description |
|---|---|
capture-tasks-from-meeting-notes | 회의록에서 Jira 태스크 자동 생성 |
generate-status-report | Jira 이슈 기반 프로젝트 상태 리포트 생성 |
search-company-knowledge | Jira에서 내부 개념/프로세스 검색 |
triage-issue | 버그 리포트 트리아지 및 중복 검색 |
spec-to-backlog | Confluence 스펙 문서를 Jira 백로그로 변환 |
# Run all BATS tests
bats tests/
# Run pre-commit hooks manually
pre-commit run --all-files
이 프로젝트는 semantic-release와 Conventional Commits를 사용하여 자동으로 버전을 관리합니다.
# Interactive commit (recommended)
bun run commit
# Or write manually
git commit -m "type(scope): description"
Types:
feat: New feature (minor version bump)fix: Bug fix (patch version bump)docs, style, refactor, test, build, ci, chore, perf: Patch version bumpplugin.json and marketplace.json are updatedpre-commit run --all-files
Validations:
Note: Pre-commit failures cannot be bypassed with
--no-verify(enforced by git-guard).
bun run commit for interactive commit creationMIT License - see LICENSE file.
Battle-tested Claude Code plugin for engineering teams — 38 agents, 156 skills, 72 legacy command shims, production-ready hooks, and selective install workflows evolved through continuous real-world use
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Claude Code skills for Godot 4.x game development - GDScript patterns, interactive MCP workflows, scene design, and shaders
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
AI-powered development tools for code review, research, design, and workflow automation.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.