security-scan

安全扫描

对整个技术栈进行全面安全扫描，包括应用程序、基础设施、容器、依赖项和云环境。此技能将基于大模型的安全分析与行业标准工具相结合，用于识别漏洞、配置错误和安全弱点。

何时使用

在以下情况下使用此技能：

• 需要对应用程序和基础设施进行全面安全评估
• 想要将多个安全扫描工具集成到统一的工作流中
• 需要借助大模型分析来识别复杂的安全问题
• 正在准备安全审计或合规认证
• 想要在 CI/CD 管道中建立基础安全扫描
• 需要扫描多个环境（云、容器、基础设施）

功能

• 基于大模型的安全分析：利用 AI 识别复杂的安全模式、业务逻辑缺陷和新型漏洞
• 集成工具生态：协调 OWASP ZAP、Snyk、Trivy、Nessus 等安全扫描器
• 多层扫描：应用程序（SAST/DAST）、基础设施（IaC 扫描）、容器、依赖项、云配置
• 漏洞关联分析：关联不同扫描工具的发现结果，以确定关键问题的优先级
• 合规映射：将漏洞映射到合规框架（SOC 2、ISO 27001、HIPAA、GDPR）
• 修复指导：为每项发现提供具体、可执行的修复步骤
• 风险评分：根据 CVSS、漏洞利用可用性和业务影响计算风险分数

示例

# Run comprehensive security scan
./scripts/security-scan.sh --target https://app.example.com

# Scan Docker containers
./scripts/security-scan.sh --container myapp:latest

# Scan infrastructure as code
./scripts/security-scan.sh --iac terraform/

# Generate compliance report
./scripts/security-scan.sh --compliance soc2

# LLM-powered security analysis
./scripts/security-scan.sh --llm-analysis --context "Payment processing system"

输出格式

Security Scan Report
─────────────────────────────────────
Scan Date: 2025-01-15T10:30:00Z
Target: https://app.example.com
Scan Duration: 2m 45s

CRITICAL FINDINGS (3):
────────────────────────
❌ SQL Injection in /api/users endpoint
   Risk: Critical (CVSS 9.8)
   Detection: OWASP ZAP + LLM analysis
   Remediation: Use parameterized queries, implement input validation
   Compliance Impact: PCI DSS 6.5.1, OWASP A1

❌ Hard-coded AWS credentials in config file
   Risk: Critical (CVSS 8.9)
   Detection: TruffleHog + LLM pattern matching
   Remediation: Move to AWS Secrets Manager, rotate credentials
   Compliance Impact: SOC 2 CC6.1, ISO 27001 A.9.4.1

❌ Unpatched vulnerability in nginx:1.18 (CVE-2021-23017)
   Risk: Critical (CVSS 9.1)
   Detection: Trivy container scan
   Remediation: Upgrade to nginx 1.20+, apply security patches
   Compliance Impact: PCI DSS 6.2, ISO 27001 A.12.6.1

HIGH FINDINGS (8):
───────────────────
⚠️ Missing Content Security Policy header
⚠️ Excessive permissions in IAM role (AdminAccess)
⚠️ Outdated OpenSSL library (CVE-2022-2068)
⚠️ Docker container running as root
⚠️ API endpoint without rate limiting
⚠️ Sensitive data in application logs
⚠️ Missing MFA for administrative access
⚠️ Unencrypted S3 bucket

MEDIUM/LOW FINDINGS (14):
──────────────────────────
ℹ️ Security headers missing (X-Frame-Options, X-Content-Type-Options)
ℹ️ Verbose error messages revealing system information
ℹ️ Session timeout too long (24 hours)
ℹ️ Cross-site request forgery (CSRF) protection missing

LLM SECURITY ANALYSIS:
──────────────────────
🔍 Business Logic Vulnerabilities:
   • Payment amount manipulation possible in checkout flow
   • Privilege escalation via IDOR in admin panel
   • Race condition in inventory reservation system

🔍 Architectural Security Issues:
   • Monolithic architecture increases attack surface
   • Lack of network segmentation between tiers
   • Insufficient logging for security events

🔍 Compliance Gaps:
   • Missing data retention policy implementation
   • Inadequate incident response procedures
   • Insufficient employee security training documentation

SUMMARY:
────────
Total Findings: 25
Critical: 3 | High: 8 | Medium: 9 | Low: 5
Risk Score: 78/100 (High Risk)
Compliance Status: 65% compliant with SOC 2

RECOMMENDATIONS:
────────────────
1. IMMEDIATE ACTION: Fix 3 critical vulnerabilities within 24 hours
2. PRIORITY: Address 8 high-risk issues within 7 days
3. IMPROVEMENTS: Implement security controls for medium/low issues
4. ARCHITECTURAL: Consider microservices segmentation, zero-trust network
5. PROCESS: Establish security training program, incident response plan

备注

• 可与现有 CI/CD 管道和安全工具集成
• 大模型分析需要仔细验证，以避免误报
• 不同的扫描工具可能有不同的许可要求
• 部分扫描器需要身份验证令牌或 API 密钥
• 采取修复措施前，务必验证发现结果
• 考虑在非高峰时段运行扫描，以尽量减少对性能的影响
• 建议对生产系统定期扫描（每天/每周）
• 保持扫描工具更新，以便检测最新漏洞