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/verify-loopgit 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/verify-loop)<a href="https://agentmods.dev/commands/sangrokjung/claude-forge/verify-loop"><img src="https://agentmods.dev/badge/commands/sangrokjung/claude-forge/verify-loop.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 | $0.00021 | $0.01071 |
| Opus 5 | $0.00010 | $0.00535 |
| Sonnet 5 | $0.00004 | $0.00214 |
| Haiku 4.5 | $0.00002 | $0.00107 |
Grade A, and why
verify-loop scanned grade A with 0 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 5d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Task
0단계: 설정 파싱
--max-retries N: 최대 재시도 횟수 (기본: 3)--only [type]: 특정 검증만 실행- 나머지: 의도 설명
1단계: 초기 환경 수집
git status --short- 변경사항 확인 (없으면 중단)git diff --name-only- 변경 파일 목록- Read:
.claude/handoff.md(있으면) - Read:
CLAUDE.md,spec.md,prompt_plan.md(있는 것만)
2단계: 의도 파악
- handoff.md 있으면 → handoff.md 기반
- $ARGUMENTS에 의도 있으면 → $ARGUMENTS 기반
- 둘 다 없으면 → 안내 후 중단:
⚠️ 의도를 알 수 없습니다. /verify-loop "변경 의도 설명"으로 재시도하세요.
3단계: 검증 루프 시작
════════════════════════════════════════════════════════════════
🔄 Verification Loop 시작 (max_retries: [N])
════════════════════════════════════════════════════════════════
각 시도마다:
[Attempt X/N]
-
코드 리뷰 (think hard):
git diff실행- 의도대로 구현됐는지
- 로직 오류, 엣지 케이스
- 불필요한 코드 (console.log, dead code)
- 보안 취약점
-
자동화 검증 (프로젝트 타입별):
- Node.js:
npm run build && npm test && npm run lint - Python:
python -m pytest && python -m flake8 - Go:
go build ./... && go test ./... - Rust:
cargo build && cargo test
- Node.js:
-
결과 출력:
├── Build: ✅/❌ ├── Test: ✅/❌ (N errors) ├── Lint: ✅/⚠️ (N fixable) └── TypeCheck: ✅/❌
4단계: 실패 시 에러 분석
🔍 에러 분석 중...
├── Fixable: N개 (타입)
└── Manual: N개 (타입)
Fixable 에러 (자동 수정):
- import 누락 → 자동 추가
- 린트 포맷 →
eslint --fix또는prettier - 미사용 변수 → 삭제 또는
_prefix - 타입 단순 오류 → 타입 추론 수정
🔧 자동 수정 중...
├── [수정 내용]
└── 완료
Manual 에러 (안내만):
⚠️ 수동 수정 필요:
1. [파일:라인] - [에러 메시지]
2. ...
💡 힌트: [해결 제안]
5단계: 재검증 또는 종료
재시도 가능하면:
- Fixable 수정 후 → 다음 Attempt로
max_retries 도달 시:
════════════════════════════════════════════════════════════════
❌ Verification Loop 실패 (N회 시도 모두 실패)
════════════════════════════════════════════════════════════════
반복 실패 에러:
1. [에러 상세]
2. ...
권장 조치:
1. /handoff 실행 - 현재 상태 기록
2. /clear 후 새 시각으로 접근
3. /learn --from-error로 교훈 기록
🎓 자동 학습 트리거: CLAUDE.md에 패턴 기록 권장
════════════════════════════════════════════════════════════════
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.
- 5d ago First seen · 129 lines · 21 tokens per session scan A 55c7bfe22090
verify-loop is a command published in the GitHub repository sangrokjung/claude-forge (822 stars, last pushed yesterday), licensed MIT. It adds 21 tokens to every session and 1,071 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
version
Display current guide and Claude Code versions.
diff-docs
Compare official Anthropic docs baseline vs current snapshot (no network — instant).
analyst
Use when performing local analyst review before pushing PR changes. Assesses code quality, impact analysis, and maintainability.
docs-review
Phase 4 of documenting-projects: Quality gate with 8 measurable criteria and iteration. Triggers: '/docs-review', invoked by documenting-projects orchestrator.
review-a11y
Review the current branch for accessibility (WCAG 2.1 AA) and responsive behavior in a React/Next.js codebase. Supports whole-branch review or scoped review by path, feature, or base branch.
review-arch
Review the current branch for code quality, TypeScript, and React/Next.js patterns. Supports whole-branch review or scoped review by path, feature, or base branch.