Agent Skills › MaxMiksa/Auto-Company › security-audit

security-audit

GitHub

用于系统性地审查应用代码、依赖项和配置的安全性，识别OWASP漏洞、CVE及配置风险。

.claude/skills/security-audit/SKILL.md MaxMiksa/Auto-Company

触发场景

审查代码安全性审计依赖项CVE 检查配置或密钥安全

安装

npx skills add MaxMiksa/Auto-Company --skill security-audit -g -y

更多选项

非标准路径

npx skills add https://github.com/MaxMiksa/Auto-Company/tree/main/.claude/skills/security-audit -g -y

不安装直接使用

npx skills use MaxMiksa/Auto-Company@security-audit

指定 Agent (Claude Code)

npx skills add MaxMiksa/Auto-Company --skill security-audit -a claude-code -g -y

安装 repo 全部 skill

npx skills add MaxMiksa/Auto-Company --all -g -y

预览 repo 内 skill

npx skills add MaxMiksa/Auto-Company --list

SKILL.md

Frontmatter

{
    "name": "security-audit",
    "agents": [
        "yokay-security-scanner"
    ],
    "description": "Use when reviewing code security, auditing dependencies for CVEs, checking configuration or secret security, assessing authentication and authorization patterns, identifying OWASP vulnerabilities (injection, XSS, CSRF), or addressing security concerns about implementations."
}

Security Audit

Systematic security review for application code, dependencies, and configuration.

Not a replacement for professional penetration testing. Identifies common vulnerabilities within scope of code review.

Audit Types

Type	Focus	When to Use
Code Review	OWASP Top 10, injection, auth	New features, PRs, suspicious code
Dependency	CVEs, outdated packages	Before deploy, periodic, CI/CD
Configuration	Secrets, permissions, hardening	Infrastructure changes, new envs
Architecture	Attack surface, data flow	Design phase, major refactors
API Security	Auth, authz, rate limiting	New endpoints, public APIs

When NOT to Use

Designing new auth flows — Use api-design for designing OAuth2/JWT endpoints from scratch
Performance issues — Use performance-optimization even if caused by auth overhead
CI/CD pipeline security — Use ci-cd for pipeline hardening (secret management, permissions)

Key Principles

Scope first — Define audit area, depth, and constraints before scanning
Classify severity — Critical (24-48h), High (1 week), Medium (2-4 weeks), Low (backlog)
Remediate or track — Fix critical issues immediately, create ohno tasks for the rest
No secrets in code — Scan for hardcoded credentials, API keys, connection strings

Quick Start Checklist

Define audit scope and type (code, dependency, config, architecture, API)
Run automated scans (npm audit, grep patterns, secret detection)
Review findings and classify severity using decision tree in references
Remediate critical/high findings immediately
Create ohno tasks for medium/low findings with appropriate priority
Document findings in audit report

References

Reference	Description
owasp-top-10.md	OWASP vulnerabilities with detection and fixes
dependency-security.md	npm audit, pip-audit, Snyk, CI/CD integration
auth-patterns.md	Secure authentication and authorization patterns
api-security.md	API-specific security concerns
secrets-management.md	Handling sensitive configuration