review

review is a command for Claude Code from studioKjm/ai-harness-template. It costs 31 tokens per session (655 once invoked), scanned A, original, MIT.

A quick code-checking command for use during implementation. It runs mechanical checks for layer rules, specification matches, and detected violations, then compares expected entities with names found in the source code.

In plain words
What is it for?
Use it between implementation steps, when drift is suspected, or before a full evaluation. You can check all gates or only layers or specification rules.
Why use it?
It catches drift, meaning the implementation is starting to differ from the planned requirements, before the whole task is finished. This gives faster feedback than a full evaluation.

Command for Claude Code

Written for Claude Code: argument-hint in frontmatter.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is bash .harness/gates/check-layers.sh.

Part of the harness plugin — 44 commands, 11 agents shipped together

Good fit Use it between implementation steps, when drift is suspected, or before a full evaluation. You can check all gates or only layers or specification rules.

Compare 6 commands from other repositories ↓
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/studioKjm/ai-harness-template
agentmods
npx agentmods add commands/studiokjm/ai-harness-template/review

Made for: Claude Code.

Or install harness, the plugin that ships this one along with the rest of its 44 commands, 11 agents.

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 review

README.md
[![agentmods](https://agentmods.dev/badge/commands/studiokjm/ai-harness-template/review.svg)](https://agentmods.dev/commands/studiokjm/ai-harness-template/review)
Your own site
<a href="https://agentmods.dev/commands/studiokjm/ai-harness-template/review"><img src="https://agentmods.dev/badge/commands/studiokjm/ai-harness-template/review.svg" alt="Measured on agentmods" height="20"></a>
Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 655 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00031 $0.00655
Opus 5 $0.00015 $0.00328
Sonnet 5 $0.00006 $0.00131
Haiku 4.5 $0.00003 $0.00065

Measured 8d ago against content hash cb7fa11f0fb5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

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 8d 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.

commands/review.md · 80 lines

What it actually says

/review — 경량 중간 검증

/evaluate의 Stage 1 (Mechanical)만 빠르게 실행한다. /run 중간에 드리프트를 조기 발견하는 용도.

When to use

  • /run 중 AC 3개 구현마다 (자동 또는 수동)
  • 드리프트가 의심될 때
  • /evaluate 전 빠른 사전 체크

Instructions

Step 1: Gate 실행 (Mechanical only)

다음 게이트만 실행한다 (/evaluate Stage 1과 동일):

# 레이어 위반 검사
bash .harness/gates/check-layers.sh

# 스펙 정합성 검사
bash .harness/gates/check-spec.sh

# 위반 탐지
bash .harness/gates/detect-violations.sh

특정 게이트만 실행 가능:

  • /review layers → check-layers.sh만
  • /review spec → check-spec.sh만
  • /review (인자 없음) → 3개 전부

Step 2: 간이 드리프트 측정

seed ontology의 핵심 엔티티가 코드에 존재하는지 확인:

  1. .harness/ouroboros/seeds/seed-v*.yaml에서 ontology 엔티티 목록 추출
  2. src/ 디렉토리에서 해당 엔티티명 검색
  3. 매칭률 계산:
    • 90%+ → OK
    • 70~90% → WARN (일부 엔티티 미구현 — 진행 단계에 따라 정상일 수 있음)
    • 70% 미만 → DRIFT (확인 필요)

Step 3: 결과 요약

═══ /review ═══════════════════════════════════════

Gates:
  check-layers:      PASS | FAIL
  check-spec:        PASS | FAIL
  detect-violations: PASS | FAIL ({n} violations)

Drift (간이):
  Ontology match: {n}/{total} entities ({percentage}%)
  Status: OK | WARN | DRIFT

Verdict: PASS | WARN | FAIL
  → PASS: 계속 진행
  → WARN: 확인 후 계속 가능
  → FAIL: 즉시 수정 필요

Time: {elapsed}
═══════════════════════════════════════════════════

/evaluate와의 차이

/review /evaluate
Stage 1 (Mechanical) O O
Stage 2 (Semantic) 간이 드리프트만 전체 AC 준수 검증
Stage 3 (Judgment) X LLM 리뷰
소요 시간 1~3분 5~15분
용도 중간 점검 최종 검증
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. 8d ago First seen · 80 lines · 31 tokens per session scan A cb7fa11f0fb5

Subscribe to this mod's changes

review is a command published in the GitHub repository studioKjm/ai-harness-template (43 stars, last pushed 3mo ago), licensed MIT. It adds 31 tokens to every session and 655 once invoked, about $0.0002 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.