| name | security-stride-methodology |
| description | Activate when conducting security analysis using STRIDE threat modeling, vulnerability assessment, and security architecture evaluation |
| license | MIT |
STRIDE Security Methodology
Guide for comprehensive security analysis using the STRIDE threat modeling framework, vulnerability assessment, and security architecture evaluation.
When to Use This Skill
Activate when:
- Conducting security threat modeling
- Performing vulnerability assessments
- Evaluating security architecture
- Creating security review documentation
- Generating threat model diagrams
- Assessing security posture of applications
STRIDE Framework
STRIDE categorizes security threats into six categories:
S - Spoofing
Identity theft/impersonation threats
- User authentication bypass
- Session hijacking
- Credential theft
- Certificate spoofing
T - Tampering
Data modification threats
- SQL injection
- Parameter manipulation
- Man-in-the-middle attacks
- Data store manipulation
R - Repudiation
Deniability threats
- Log tampering
- Missing audit trails
- Insufficient logging
- Non-repudiation failures
I - Information Disclosure
Data exposure threats
- Sensitive data leakage
- Error message exposure
- Directory traversal
- Improper access controls
D - Denial of Service
Availability threats
- Resource exhaustion
- Rate limiting bypass
- Memory leaks
- Infinite loops
E - Elevation of Privilege
Authorization bypass threats
- Privilege escalation
- Role bypass
- Insecure direct object reference
- Broken access control
Security Analysis Process
Phase 1: Asset Identification
- Identify critical data assets
- Map data flows and trust boundaries
- Identify entry points and exit points
- Document authentication mechanisms
- List third-party dependencies
Phase 2: Threat Modeling
- Apply STRIDE to each component
- Identify threat scenarios
- Assess likelihood and impact
- Document attack vectors
- Create threat diagrams
Phase 3: Vulnerability Assessment
- Static code analysis
- Dependency vulnerability scanning
- Configuration review
- Authentication/authorization testing
- Input validation analysis
Phase 4: Risk Prioritization
- Calculate risk scores
- Prioritize by business impact
- Identify quick wins
- Create remediation roadmap
- Document mitigations
Threat Model Template
## Threat: [STRIDE Category] - [Threat Name]
### Description
[Detailed threat description]
### Attack Vector
[How the attack could be executed]
### Affected Components
- [Component 1]
- [Component 2]
### Risk Assessment
- **Likelihood**: High/Medium/Low
- **Impact**: High/Medium/Low
- **Risk Score**: [Likelihood × Impact]
### Mitigation
- [Mitigation step 1]
- [Mitigation step 2]
### Verification
- [How to verify mitigation is effective]
Mermaid Diagram Patterns
Data Flow Diagram
graph LR
A[User] -->|HTTPS| B[Load Balancer]
B -->|Internal| C[App Server]
C -->|Query| D[(Database)]
C -->|API| E[External Service]
style B fill:#f9f,stroke:#333
style D fill:#bbf,stroke:#333
Trust Boundary Diagram
graph TB
subgraph "Untrusted Zone"
A[Internet Users]
end
subgraph "DMZ"
B[Web Server]
end
subgraph "Trusted Zone"
C[App Server]
D[(Database)]
end
A --> B
B --> C
C --> D
Security Report Structure
# Security Assessment Report
## Executive Summary
- **Overall Risk Level**: [Critical/High/Medium/Low]
- **Critical Issues**: [Count]
- **High Issues**: [Count]
- **Recommendations**: [Top 3 actions]
## Threat Model Analysis
### STRIDE Coverage
- Spoofing: [Finding count]
- Tampering: [Finding count]
- Repudiation: [Finding count]
- Information Disclosure: [Finding count]
- Denial of Service: [Finding count]
- Elevation of Privilege: [Finding count]
## Architecture Diagram
[Mermaid diagram with trust boundaries]
## Findings
### Critical
1. [Finding with STRIDE category]
### High Priority
1. [Finding with STRIDE category]
## Remediation Roadmap
### Immediate (0-7 days)
- [Action items]
### Short-term (1-4 weeks)
- [Action items]
### Long-term (1-3 months)
- [Action items]
Key Principles
- Defense in Depth: Multiple security layers
- Least Privilege: Minimal access rights
- Secure by Default: Safe default configurations
- Fail Securely: Handle errors safely
- Zero Trust: Verify everything, trust nothing
- Data Classification: Protect based on sensitivity