What problem does it solve? Security scans need a consistent, repository-wide threat model to ground vulnerability discovery, but teams often lack one or reuse stale, diff-scoped assumptions. This Skill creates or reuses a cached repository-scoped threat model so every scan pass reasons from the same trust boundaries and attacker stories. ## Core Features & Use Cases - Threat Model Generation: Produces a structured Markdown threat model with overview, trust boundaries, attack surfaces, mitigations, and severity calibration from Critical to Low. - Cache Reuse and Versioning: Reuses an existing threat model only when its Repository and Version lines match the current target, otherwise regenerates it. - Independent Pass Mode: Generates fresh, isolated threat models for multi-pass deep security scans without touching the shared cache. - Use Case: During a deep security scan of a web application repository, invoke this Skill to produce a threat model identifying attacker-controlled inputs, authn/authz boundaries, and which vulnerability classes matter most before any findings are reported. ## Quick Start Ask the AI to generate a repository-scoped threat model for the current repository using the security threat-model skill.
