Threat Modeling Expert
Expert in threat modeling methodologies, security architecture review, and risk assessment using STRIDE, PASTA, attack trees, and security requirement extraction.
by brandonwise · published 2026-03-22
$ claw add gh:brandonwise/brandonwise-threat-modeling# Threat Modeling Expert
Expert in threat modeling methodologies, security architecture review, and risk assessment using STRIDE, PASTA, attack trees, and security requirement extraction.
Description
USE WHEN:
DON'T USE WHEN:
---
Core Process
1. Define Scope
2. Create Data Flow Diagram
[User] → [Web App] → [API Gateway] → [Backend] → [Database]
↓
[External API]3. Identify Assets & Entry Points
4. Apply STRIDE
5. Build Attack Trees
Goal: Access Admin Panel
├── Steal admin credentials
│ ├── Phishing
│ ├── Brute force
│ └── Session hijacking
├── Exploit vulnerability
│ ├── SQL injection
│ └── Auth bypass
└── Social engineering
└── Support desk compromise6. Score & Prioritize
Use DREAD or CVSS:
7. Design Mitigations
Map threats to controls and validate coverage.
8. Document Residual Risks
What's accepted vs. mitigated.
---
STRIDE Analysis Template
| Component | Spoofing | Tampering | Repudiation | Info Disclosure | DoS | EoP |
|-----------|----------|-----------|-------------|-----------------|-----|-----|
| Web App | Auth bypass | XSS, CSRF | Missing logs | Error messages | Rate limit | Broken access |
| API | Token theft | Input manip | No audit | Data exposure | Resource exhaust | Privilege escalation |
| Database | Credential theft | SQL injection | No audit trail | Backup exposure | Connection flood | Direct access |
---
Threat Categories by Layer
Application Layer
Network Layer
Infrastructure Layer
Human Layer
---
Data Flow Diagram Elements
| Element | Symbol | Description |
|---------|--------|-------------|
| External Entity | Rectangle | Users, external systems |
| Process | Circle | Application logic |
| Data Store | Parallel lines | Database, cache, files |
| Data Flow | Arrow | Data movement |
| Trust Boundary | Dashed line | Security perimeter |
---
Risk Prioritization Matrix
LOW IMPACT HIGH IMPACT
HIGH LIKELIHOOD MEDIUM HIGH
LOW LIKELIHOOD LOW MEDIUMDREAD Scoring (1-10 each)
| Factor | Question |
|--------|----------|
| Damage | How bad if exploited? |
| Reproducibility | How easy to reproduce? |
| Exploitability | How easy to attack? |
| Affected Users | How many impacted? |
| Discoverability | How easy to find? |
**Score**: Sum / 5 = Risk Level
---
Mitigation Strategies
Input Validation
Authentication
Authorization
Cryptography
Monitoring
---
Best Practices
1. **Involve developers** in threat modeling sessions
2. **Focus on data flows**, not just components
3. **Consider insider threats**
4. **Update models** with architecture changes
5. **Link threats** to security requirements
6. **Track mitigations** to implementation
7. **Review regularly**, not just at design time
8. **Keep models living documents**
---
Output Template
# Threat Model: [System Name]
## Scope
- Components in scope
- Out of scope
## Assets
- Critical assets list
## Trust Boundaries
- Internal vs external
- Admin vs user
## Data Flow Diagram
[DFD here]
## STRIDE Analysis
[Table here]
## Prioritized Threats
1. [High] Description - Mitigation
2. [Medium] Description - Mitigation
## Residual Risks
- Accepted risks with justification
## Review Schedule
- Next review dateMore tools from the same signal band
Order food/drinks (点餐) on an Android device paired as an OpenClaw node. Uses in-app menu and cart; add goods, view cart, submit order (demo, no real payment).
Sign plugins, rotate agent credentials without losing identity, and publicly attest to plugin behavior with verifiable claims and authenticated transfers.
The philosophical layer for AI agents. Maps behavior to Spinoza's 48 affects, calculates persistence scores, and generates geometric self-reports. Give your...