Changelog
Initial release in the security skill pack.
How it works
1
Identify system
Checks arguments for feature or system name
2
Load architecture
Reads engineering context, specs, infrastructure docs from the brain
3
Map assets and boundaries
Identifies what's valuable and where trust levels change
4
Apply STRIDE
Evaluates each component for all six threat categories
5
Generate model
Risk matrix with prioritized threats and specific mitigations
6
Save
Writes to engineering/security/threat-model-<slug>.md
Details
- Reads your actual architecture from engineering/CLAUDE.md - not generic
- Covers all STRIDE categories: Spoofing, Tampering, Repudiation, Info Disclosure, DoS, Elevation
- Each threat gets a risk level, description, and specific mitigation
- Run before launching new features to think through attack vectors
- If architecture docs don't exist, asks you to describe the system