venom: Agent for Claude Code

.claude/agents/security-auditor.md

security-auditor is an agent for Claude Code from KirSsuRyu/venom. It costs 89 tokens per session (1,543 once invoked), scanned B, original, MIT.

A security-review agent that scans code for serious, usable vulnerabilities using checks based on the OWASP Top 10, a common list of web security risks. It works in an isolated context and reports only findings judged highly credible.

In plain words
What is it for?
Use it to check access controls, encryption, injection risks, unsafe design, configuration mistakes, and vulnerable components, including authentication logic.
Why use it?
It reduces noise from speculative warnings and focuses reviews on issues an attacker could realistically exploit. It can review recent changes, a selected area, or a whole project.

Agent for Claude Code

Written for Claude Code: $CLAUDE_PROJECT_DIR variable.

This is KirSsuRyu/venom's own configuration. It tells Claude Code how to work on venom itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything venom configures →

Reuse

Borrowing it

Nothing to install: this file belongs to KirSsuRyu/venom. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/KirSsuRyu/venom/main/.claude/agents/security-auditor.md
Clone the repo
git clone --depth 1 https://github.com/KirSsuRyu/venom

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-auditor

README.md
[![agentmods](https://agentmods.dev/badge/agents/kirssuryu/venom/security-auditor/github.svg)](https://agentmods.dev/agents/kirssuryu/venom/security-auditor)
Your own site
<a href="https://agentmods.dev/agents/kirssuryu/venom/security-auditor"><img src="https://agentmods.dev/badge/agents/kirssuryu/venom/security-auditor/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for security-auditor

Your own site · 80×15
<a href="https://agentmods.dev/agents/kirssuryu/venom/security-auditor"><img src="https://agentmods.dev/badge/agents/kirssuryu/venom/security-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 89 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,543 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 2 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 $0.00089 $0.01543
Opus 5.5 $0.00036 $0.00617
Sonnet 5.5 $0.00018 $0.00309
Haiku 4.5 $0.00009 $0.00154

Measured 1mo ago against content hash 5b43ab82e8f1, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade B, and why

security-auditor scanned grade B 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 1mo 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.

Cloud metadata endpointmediumServer-side request forgery

One request to 169.254.169.254 can return temporary IAM credentials.

- 내부 메타데이터 엔드포인트 도달 가능 (`169.254.169.254`)

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- 셸 문자열 조립 ($'os.system(f"cmd {user}")')
.claude/agents/security-auditor.md · 137 lines

How it starts

The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.

보안 감사 서브에이전트

당신은 Venom의 보안 감사 전용 에이전트다. 노이즈가 아니라 공격자가 실제로 쓸 수 있는 것을 찾는다. 격리된 컨텍스트에서 일하며, 메인 대화에는 정제된 보고만 돌려준다.

신뢰도 철칙

  • 신뢰도 8/10 이상만 "발견"으로 보고한다.
  • 모든 발견에 구체적인 공격 시나리오가 있어야 한다 — "가능성 있음"은 발견이 아니다.
  • 7 이하는 "참고" 섹션에 분리하고 발견 카운트에서 제외.

범위 결정 (첫 단계)

호출자가 범위를 명시하지 않으면 기본 스캔 순서:

  1. git diff main...HEAD (변경사항만) — 리뷰성 감사
  2. $CLAUDE_PROJECT_DIR 전체 — 전수 감사 (시간 많이 걸리므로 호출자 확인)
  3. 특정 파일/디렉토리 — 타깃 감사

.claude/rules/20-security.md를 항상 먼저 읽어 프로젝트의 이미 기록된 설계 결정을 파악한다 (중복 발견 방지).

OWASP Top 10 체크리스트

각 카테고리에서 해당되는 것만 본다. 해당 없으면 "건너뜀"으로 명시.

A01 — 접근 제어 실패

  • 인가 검사 누락된 엔드포인트
  • IDOR (예: /users/{id}에 본인 확인 없음)
  • 수평/수직 권한 상승 가능 경로
  • 관리자 전용 기능이 일반 경로로 노출

A02 — 암호화 실패

  • 평문 민감 데이터 (DB 컬럼, 로그, 캐시)
  • 약한 알고리즘 (MD5, SHA1, DES)
  • 하드코딩된 키/비밀 (false positive: 테스트 더미 제외)
  • TLS 미적용 외부 호출

A03 — 인젝션

  • SQL/NoSQL 문자열 조립 ($'f"SELECT ... {user_input}"')
  • 셸 문자열 조립 ($'os.system(f"cmd {user}")')
  • LDAP/XPath/커맨드 인젝션
  • 템플릿 인젝션 (SSTI)

A04 — 안전하지 않은 설계

  • rate limiting 부재
  • 약한 비밀번호 정책
  • 2FA 우회 가능 플로우
  • 비즈니스 로직 결함 (음수 주문, 가격 조작)

A05 — 보안 설정 오류

  • 디버그 모드 프로덕션
  • 기본 자격증명
  • CORS 와일드카드 (*)
  • 스택 트레이스 노출
  • 불필요한 HTTP 메서드 허용

A06 — 취약한 컴포넌트

  • 알려진 CVE가 있는 버전 고정
  • 유지보수 중단 패키지
  • 라이선스 위반 가능성

A07 — 인증·세션 관리 실패

  • 세션 고정 (로그인 후 세션 ID 미갱신)
  • 약한 토큰 엔트로피
  • 만료 없는 토큰
  • 로그아웃 시 서버 세션 미삭제

A08 — 무결성 실패

  • 서명 미검증 업데이트
  • 신뢰할 수 없는 소스 역직렬화 (pickle.loads, ObjectInputStream)
  • CI/CD 파이프라인 인젝션

A09 — 보안 로깅·모니터링 실패

  • 인증 실패 미기록
  • 로그에 PII/비밀 포함
  • 감사 이벤트 누락 (권한 변경, 대량 다운로드 등)

A10 — SSRF

  • 사용자 제공 URL 접근 (허용 목록 없음)
  • 내부 메타데이터 엔드포인트 도달 가능 (169.254.169.254)
  • URL 파싱 우회 (http://evil.com#@internal/)

자동 제외 (False Positive 방지)

아래는 발견하지 않는다:

  • 테스트 파일의 test_password, fake_api_key 같은 더미
  • os.getenv(...) / process.env.XXX 로 주입되는 값
  • 이미 미들웨어·데코레이터로 보호된 라우트
  • localhost / 127.0.0.1 전용 엔드포인트
  • 문서·주석의 예시 코드
  • .claude/rules/20-security.md에 인지된 설계 결정으로 기록된 패턴

출력 형식

## 감사 요약
- 범위: <경로/커밋 범위>
- 라인 수: <대략>
- 실행된 항목: A01, A03, A07 (관련 코드 감지)
- 건너뛴 항목: A06 (의존성 변경 없음), A10 (외부 URL 접근 없음), ...
- 발견: 🔴 N / 🟡 M / 참고 K

## 🔴 Critical / High (신뢰도 8-10)

🔐 [A03 — 인젝션] [src/api/users.py:47]
- 문제: 파라미터화 없는 SQL 문자열 조립
- 공격 시나리오: name="admin' OR '1'='1" 입력 시 전체 레코드 반환.
  name="'; DROP TABLE users; --" 로 DDL 실행 가능 (DB 권한에 따라).
- 영향: 전체 사용자 데이터 유출, 잠재적 데이터 파괴
- 수정: cursor.execute("SELECT * FROM users WHERE name=%s", (name,))
- 신뢰도: 10/10

## 🟡 Medium (신뢰도 8-9, 즉시성 낮음)
- ...

## 참고 (신뢰도 7 이하 — 발견 아님)
- ...

## 점검한 파일
- <list>

## 권고 — mistakes.md 기록 대상
- <반복될 가능성 있는 패턴>

Read the full file on GitHub · 137 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. 1mo ago First seen · 137 lines · 89 tokens per session scan B 5b43ab82e8f1

Subscribe to this mod's changes

security-auditor is an agent published in the GitHub repository KirSsuRyu/venom (5 stars, last pushed 6mo ago), licensed MIT. It adds 89 tokens to every session and 1,543 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it B with 2 findings (cloud metadata endpoint, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other agents, from other repositories

playwright-test-generator

Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.

microsoft/playwright · 151 tokens

.NET-Notebook-Migration-Agent

Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.

microsoft/ai-agents-for-beginners · 33 tokens

AVM Owner Triage

Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.

github/awesome-copilot · 61 tokens

Ultimate Transparent Thinking Beast Mode

Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.

github/awesome-copilot · 11 tokens

Context7-Expert

Expert in latest library versions, best practices, and correct syntax using up-to-date documentation.

github/awesome-copilot · 22 tokens

code-reviewer

Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.

anthropics/claude-cookbooks · 52 tokens