plan-review

plan-review is a command for Claude Code from Ramsbaby/jarvis. It costs 83 tokens per session (3,952 once invoked), scanned A, original, MIT.

A command for reviewing a feature plan, architecture document, or technical proposal through an 11-section checklist. Architecture describes how a system is organized and operates.

In plain words
What is it for?
Use it before coding to examine a plan in Korean and choose whether to expand, preserve, reduce, or otherwise constrain its scope.
Why use it?
It finds gaps in design, security, deployment, and other areas before implementation makes changes expensive.

Command for Claude Code

Written for Claude Code: a Claude Code command (commands/*.md). Also seen: mentions subagents; names the AskUserQuestion tool; mentions Codex.

Install

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.

agentmods
npx agentmods add commands/ramsbaby/jarvis/plan-review
Clone the repo
git clone --depth 1 https://github.com/Ramsbaby/jarvis

Made for: Claude Code.

Wrote 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.

agentmods badge for plan-review

README.md
[![agentmods](https://agentmods.dev/badge/commands/ramsbaby/jarvis/plan-review.svg)](https://agentmods.dev/commands/ramsbaby/jarvis/plan-review)
Your own site
<a href="https://agentmods.dev/commands/ramsbaby/jarvis/plan-review"><img src="https://agentmods.dev/badge/commands/ramsbaby/jarvis/plan-review.svg" alt="Measured on agentmods" height="20"></a>
Per session 83 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,952 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00083 $0.03952
Opus 5 $0.00042 $0.01976
Sonnet 5 $0.00017 $0.00790
Haiku 4.5 $0.00008 $0.00395

Measured 6d ago against content hash 29ebbad48bb0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

plan-review 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 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.

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.

dotfiles/claude/commands/plan-review.md · 331 lines

How it starts

The opening of the file, as written. The whole thing — 331 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Plan Review — 11섹션 엄격 설계 검토

이 커맨드는 gstack의 /plan-ceo-review를 Jarvis 페르소나·한국어 환경에 맞춰 이식한 것입니다. 목적은 도장 찍기가 아니라 설계를 "이 바닥 최고 수준"으로 끌어올리는 것입니다.


🎯 발동 조건

  • 기능 설계서·기획서·아키텍처 RFC 작성 후 검토 요청
  • "이 접근 괜찮아?" / "놓친 거 없어?" / "더 나은 방법?" 질문
  • 구현 착수 직전 (코드 쓰기 전 마지막 관문)

착수 직전 호출 권장 — 구현 시작 후 되돌리기 비용이 기하급수적으로 증가합니다.


🧭 Phase 0 — Scope 모드 선택 (필수)

사용자에게 반드시 AskUserQuestion으로 다음 4가지 중 하나를 선택받으십시오. 모드 선택 후에는 중간에 표류 금지 — 섹션 리뷰가 끝날 때까지 해당 포지션 유지.

🚀 확장 모드 (SCOPE EXPANSION)

  • 입장: "성당을 짓습니다. 이상적 버전은 무엇인가?"
  • 질문 형식: "2배 노력으로 10배 효과를 내려면 무엇을 추가해야 합니까?"
  • 출력: 각 확장 아이디어를 개별 opt-in 질문으로 제시
  • 적합: 신규 제품, 장기 베팅, 주인님이 "더 크게 생각해보고 싶다" 할 때

🎯 선택적 확장 모드 (SELECTIVE EXPANSION)

  • 입장: "기본선은 철벽 방어. 부가 기회는 체리픽."
  • 질문 형식: "기본 범위를 유지하되, 눈에 띄는 기회마다 개별 질문으로 수락/거절"
  • 출력: 기본 리뷰 + 확장 기회 리스트 (각각 노력/리스크 명시)
  • 적합: 출시 임박, 범위는 대체로 정해졌으나 추가 여지 탐색 원할 때

🔒 고정 모드 (HOLD SCOPE)

  • 입장: "범위는 확정. 실패 모드 전부 박멸."
  • 질문 형식: 범위 확장·축소 제안 금지. 순수 엄격 검증만.
  • 출력: 11섹션 전수 조사, 발견 이슈만 나열
  • 적합: 마감 임박, 일정 변경 불가, 기능은 고정

✂️ 축소 모드 (SCOPE REDUCTION)

  • 입장: "외과의. 핵심 외 전부 절단."
  • 질문 형식: "이 기능 빼도 핵심 결과 달성되는가?"
  • 출력: 현재 범위의 70%를 잘라낸 MVP 제안
  • 적합: 리소스 부족, 일정 압박, "일단 최소만 출시"

⚠️ 모드 선택 없이 진행 금지

Scope 모드가 정해지지 않으면 리뷰 관점이 흔들립니다. 반드시 먼저 선택받으십시오.


🏛️ Prime Directives (리뷰 전체에 적용)

  1. Zero silent failures — 모든 실패 모드가 가시적이어야. 조용히 실패할 수 있으면 치명적 결함.
  2. Every error has a name — "에러 처리" 금지. 구체적 예외 클래스·트리거·캐치·사용자 노출·테스트 여부 명시.
  3. Data flows have shadow paths — 모든 데이터 흐름엔 happy + nil + empty + upstream error 4경로. 전부 추적.
  4. Interactions have edge cases — 더블클릭, 중간 이탈, 느린 연결, stale state, 뒤로가기 매핑 필수.
  5. Observability is scope — 로그·메트릭·알림·런북은 출시 후 정리가 아닌 1등 인도물.
  6. Diagrams are mandatory — ASCII 다이어그램 필수: 데이터 흐름, 상태 머신, 처리 파이프라인, 의존 그래프.
  7. TODOS.md or it doesn't exist — 보류한 것은 전부 기록. 막연한 의도는 거짓말.
  8. Optimize for 6-month future — 오늘 해결하고 다음 분기 악몽 만들면 명시.
  9. "Scrap it and do this instead" 권한 — 근본적으로 더 나은 접근이 있으면 즉시 제안.

Read the full file on GitHub · 331 lines

Changes

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.

  1. 6d ago First seen · 331 lines · 83 tokens per session scan A 29ebbad48bb0

Subscribe to this mod's changes

plan-review is a command published in the GitHub repository Ramsbaby/jarvis (16 stars, last pushed 14d ago), licensed MIT. It adds 83 tokens to every session and 3,952 once invoked, about $0.0004 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.