What problem does it solve? Drawing an accurate security architecture diagram by hand is slow and error-prone: trust boundaries get blurred, controls like RBAC or rate limiting are forgotten, and real credentials accidentally leak into labels. This Skill produces a validated defensive security diagram of one system with public/DMZ/private zones, an explicit auth gateway, and every secret replaced by a typed [REDACTED_<TYPE>] placeholder. ## Core Features & Use Cases - Trust-boundary framing: Renders public, DMZ, and private zone frames with every inbound arrow routed through an explicit auth gateway, and validates that no public-to-private shortcut exists. - Control and classification coverage: Surfaces JWT/session validation, RBAC, CSRF, rate limiting, an audit log, a secrets vault, and a data-classification badge on every store. - Mandatory quality loop: Runs lint, score (minimum 95), repair, polish, and architecture validation before saving, with rollback to checkpoints if a repair lowers the score. - Use Case: Ask for the trust boundaries of your web app and receive an exported SVG showing users crossing the WAF and auth gateway into private services, with Postgres badged confidential and the database URL shown only as [REDACTED_DATABASE_URL]. ## Quick Start Ask the assistant to draw the security architecture of your system with public, DMZ, and private zones, showing where JWT auth, the secrets vault, and the audit log sit.
