Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/sangrokjung/claude-forge/worktree-startgit clone --depth 1 https://github.com/sangrokjung/claude-forgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/sangrokjung/claude-forge/worktree-start)<a href="https://agentmods.dev/commands/sangrokjung/claude-forge/worktree-start"><img src="https://agentmods.dev/badge/commands/sangrokjung/claude-forge/worktree-start.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00020 | $0.01739 |
| Opus 5 | $0.00010 | $0.00870 |
| Sonnet 5 | $0.00004 | $0.00348 |
| Haiku 4.5 | $0.00002 | $0.00174 |
Grade D, and why
worktree-start scanned grade D with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
grep -E "\.claude/context/|\.claude/settings" .gitignore 2>/dev/null Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
2. 삭제 후 재생성: rm -rf [경로] && /worktree-start [브랜치] How it starts
The opening of the file, as written. The whole thing — 227 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Task
0단계: 환경 확인
git rev-parse --git-dir
git worktree list
git branch -a
git 레포 아니면:
❌ git 레포가 아닙니다.
git init 먼저 실행하세요.
→ 중단
0.5단계: 충돌 방지 확인
# .gitignore에 Claude Code 자동 생성 파일이 있는지 확인
grep -E "\.claude/context/|\.claude/settings" .gitignore 2>/dev/null
# 추적 중인 Claude Code 파일 확인
git ls-files | grep "^\.claude/"
Claude Code 파일이 .gitignore에 없으면:
⚠️ 충돌 방지 설정 필요
════════════════════════════════════════
병렬 워크트리에서 Claude Code 자동 생성 파일로 인한
병합 충돌이 발생할 수 있습니다.
.gitignore에 추가 필요:
.claude/context/
.claude/settings.json
.claude/settings.local.json
자동 추가 명령어:
cat >> .gitignore << 'EOF'
# Claude Code 자동 생성 파일 (충돌 방지)
.claude/context/
.claude/settings.json
.claude/settings.local.json
EOF
git add .gitignore
git commit -m "chore: .gitignore에 Claude Code 자동 생성 파일 추가"
계속 진행하시겠습니까? [y/N]
════════════════════════════════════════
→ 사용자 응답 대기
Claude Code 파일이 Git에 추적 중이면:
⚠️ Claude Code 파일이 Git에 추적 중입니다
════════════════════════════════════════
다음 파일이 추적 중이어서 병합 충돌이 발생할 수 있습니다:
[git ls-files 결과]
추적 해제 명령어:
git rm -r --cached .claude/context/
git add .gitignore
git commit -m "chore: Claude Code 자동 생성 파일 추적 해제"
계속 진행하시겠습니까? [y/N]
════════════════════════════════════════
→ 사용자 응답 대기
1단계: 인자 파싱
$ARGUMENTS에서 옵션 추출:
--type feature→ 워크플로우 타입: feature (기본값)--type bugfix→ 워크플로우 타입: bugfix--type refactor→ 워크플로우 타입: refactor- 나머지 → 브랜치명
브랜치명 결정:
브랜치명 있으면:
feature/,fix/,refactor/접두사 없으면 타입에 따라 자동 추가:--type feature(기본) →feature/접두사--type bugfix→fix/접두사--type refactor→refactor/접두사
- 예:
auth --type bugfix→fix/auth
브랜치명 없으면:
❓ 브랜치명을 입력하세요.
사용법: /worktree-start auth
/worktree-start feature/payment
/worktree-start fix/login-bug
/worktree-start refactor/utils --type refactor
→ 중단
2단계: 워크트리 생성
# 현재 디렉토리명 기준 상위에 생성
# 예: /Users/dev/myproject + feature/auth → /Users/dev/myproject-auth
BRANCH_NAME="[결정된 브랜치명]"
WORKTREE_DIR="../$(basename $(pwd))-${BRANCH_NAME##*/}"
# 브랜치 존재 여부 확인
git show-ref --verify --quiet refs/heads/$BRANCH_NAME
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 227 lines · 20 tokens per session scan D f99114aa7e34
worktree-start is a command published in the GitHub repository sangrokjung/claude-forge (825 stars, last pushed 2d ago), licensed MIT. It adds 20 tokens to every session and 1,739 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it D with 2 findings (reads agent configuration directories, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
CONTEXT
Slash commands for SPEC-First development workflow.
review
Review plans with multi-angle analysis (mandatory + extended + autonomous).
review
Perform a thorough code review of changes in this branch.
01_confirm
Extract plan from conversation, create file in draft/, auto-apply non-BLOCKING improvements, move to pending.
05_cleanup
Safe dead code cleanup - detect and remove unused imports and dead files with auto-apply for Low/Medium risk.
add-tool
Scaffold and implement a new tool for VisionAgent.