threat-model

name	threat-model
description	Generate a STRIDE-based threat model from codebase and architecture analysis. Identifies assets, trust boundaries, data flows, threats, and mitigations. Language-agnostic.
license	MIT
allowed-tools	["Read","Glob","Grep","Bash","AskUserQuestion"]

Threat Model Generator

When asked to create a threat model, analyze the codebase and generate a structured threat model using the STRIDE methodology.

Step 1: System Discovery

Map the system by examining the codebase:

1. Entry points: API routes, web handlers, CLI commands, message consumers, cron jobs
2. Data stores: Databases, caches, file systems, object storage, message queues
3. External services: Third-party APIs, OAuth providers, payment gateways, email services
4. Authentication mechanisms: How users prove identity (JWT, sessions, API keys, OAuth)
5. User roles: Admin, regular user, anonymous, service accounts, internal services
6. Deployment: Docker, k8s, cloud services (check Dockerfile, docker-compose, terraform, CI configs)

Step 2: Identify Assets

List what an attacker would want:

• Data assets: User PII, credentials, payment data, business data, API keys, session tokens
• System assets: Server access, database access, admin panels, CI/CD pipelines