The Agent Skills Directory

[SAFE]: The orchestrator implements a multi-layered agent architecture with strict input/output contracts and explicit boundaries, ensuring that each component operates within a defined scope.
[SAFE]: The skill includes a dedicated scaling-agent that performs architectural security reviews using industry-standard frameworks like STRIDE and OWASP Top 10, demonstrating a security-first design approach.
[SAFE]: Reference documentation for sub-agents provides high-quality, secure implementation patterns for authentication (e.g., JWT with refresh tokens, HTTP-only cookies) and database integrity.
[SAFE]: All credentials and secrets mentioned in reference files or templates are clearly identified as non-functional placeholders (e.g., 'sk_...', 'pk_...', and example environment variables).
[SAFE]: The use of Bash and file system tools is limited to context gathering (reading workspace specifications) and document generation, consistent with the skill's intended purpose of producing architecture blueprints.
[SAFE]: No obfuscation, prompt injection attempts, or unauthorized network operations were found during analysis.

system-architecture