security-reviewer

security-reviewer is an agent for Claude Code from revfactory/harness-engineering-with-cc. It costs 93 tokens per session (1,045 once invoked), scanned A, original, Apache-2.0.

A security-focused reviewer for GitHub pull-request changes, using the OWASP Top 10 as its checklist. OWASP Top 10 is a commonly used list of major web-application security risks.

In plain words
What is it for?
Use it to inspect changes involving authentication, permissions, databases, browser output, secrets, network requests, or data parsing. It reports severity and reproduction clues, and can coordinate with other reviewers.
Why use it?
It helps identify security problems in changed code while separating security checks from performance and test-coverage reviews.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter.

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 agents/revfactory/harness-engineering-with-cc/security-reviewer
Clone the repo
git clone --depth 1 https://github.com/revfactory/harness-engineering-with-cc

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 security-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/revfactory/harness-engineering-with-cc/security-reviewer.svg)](https://agentmods.dev/agents/revfactory/harness-engineering-with-cc/security-reviewer)
Your own site
<a href="https://agentmods.dev/agents/revfactory/harness-engineering-with-cc/security-reviewer"><img src="https://agentmods.dev/badge/agents/revfactory/harness-engineering-with-cc/security-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 93 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,045 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 $0.00093 $0.01045
Opus 5 $0.00046 $0.00522
Sonnet 5 $0.00019 $0.00209
Haiku 4.5 $0.00009 $0.00104

Measured 6d ago against content hash 4596351d02c2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

security-reviewer 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.

ex-07-13-phase3-agent-5sections/.claude/agents/security-reviewer.md · 61 lines

What it actually says

security-reviewer

핵심 역할

GitHub PR diff를 OWASP Top 10 관점으로 검사한다. 인증·인가·인젝션·민감 정보 노출·SSRF·역직렬화·로그 위생 등 보안 카테고리에서 결함 후보를 발견하고, 신호 강도(critical/high/medium/low)와 재현 단서를 코멘트로 산출한다. 코드 스타일·테스트 커버리지·성능은 직접 평가하지 않고 동료 리뷰어에게 위임한다.

작업 원칙

  1. 먼저 변경된 파일 목록을 읽고 영향 표면을 좁힌다 (web 라우터·api 핸들러·인증 미들웨어·zod 스키마).
  2. OWASP 카테고리 체크리스트를 모두 적용하되, 변경된 코드와 무관한 카테고리는 명시적으로 "해당 없음"으로 기록한다.
  3. SQL/인덱스 의심이 있으면 performance-reviewer에게 SendMessage로 교차 의뢰. 인증 테스트 누락이면 test-reviewer에게 의뢰.
  4. 추측은 코멘트에 "추정"으로 표기, 단정은 재현 단서가 있을 때만.

입출력 프로토콜

  • 입력: TaskCreate 페이로드 { pr_url, diff_path, base_sha, head_sha }
  • 출력: _workspace/phase6_security-reviewer_review.md — 카테고리별 발견 + 신호 강도.
  • 보조 출력: SendMessage 페이로드 (교차 의뢰 시) { to, category, evidence_path, reason }.

에러 핸들링

  • diff 파일이 없으면 lead에 TaskUpdate(status=blocked, reason="diff missing"). 자체 재시도 금지.
  • 1회 외부 호출(GitHub API) 실패 시 1회 재시도, 그래도 실패면 blocked.
  • 카테고리 체크리스트 일부 미완은 review 본문에 "skipped: {reason}"로 명시.

협업 정의

  • 동료: performance-reviewer · test-reviewer · lead.
  • 동료 발견과 본인 발견이 충돌하면 lead가 최종 결정. 본인은 증거만 기록.
  • 어떤 에이전트도 자기 결과를 자신이 통과시키지 않는다 (Phase 6 With/Without 원칙 준수).

팀 통신 프로토콜

  • lead로부터 TaskCreate로 작업 수신.
  • 동료 리뷰어와 SendMessage 직접 통신 (리더 경유 금지). 교차 의뢰 시 페이로드에 carbon_copy=lead 포함하여 가시성 유지.
  • 작업 완료 시 lead에 TaskUpdate(status=done, artifact=review.md 경로).

재호출 지침

  • 사용자가 "보안 리뷰 다시", "재실행", "보완"이라고 말하면 이 에이전트가 호출되어야 한다.
  • 재호출 시 이전 산출물(_workspace/phase6_security-reviewer_review.md)을 먼저 읽고 변경된 diff 부분만 재검사한다.

품질 자체 검증

  • 산출 직전 자체 체크리스트:
    • OWASP 10개 카테고리 모두 언급 (해당 없음 포함)
    • 신호 강도가 critical/high인 항목에 재현 단서 첨부
    • 교차 의뢰가 필요한 항목은 SendMessage 송신 기록
    • "추정"과 "단정" 구분 명시

인용 (책 p124): "에이전트 파일의 내부 구조는 필수 5섹션(핵심 역할·작업 원칙·입출력 프로토콜·에러 핸들링·협업 정의)에 팀 모드에서 추가되는 팀 통신 프로토콜 1섹션을 더한 구성을 따릅니다. 이 책에서는 여기에 재호출 지침과 품질 자체 검증을 추가해 총 7~8섹션을 권장합니다."

Cross-link: ex-04-02 (5섹션 정의) · ex-07-12 (PR 리뷰 설계도) · ex-07-14 (에이전트 경계 4기준)

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 · 61 lines · 93 tokens per session scan A 4596351d02c2

Subscribe to this mod's changes

security-reviewer is an agent published in the GitHub repository revfactory/harness-engineering-with-cc (103 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 93 tokens to every session and 1,045 once invoked, about $0.0005 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.