The Agent Skills Directory

[SAFE]: A comprehensive audit of the skill instructions and reference materials revealed no malicious behavior, obfuscation, or security vulnerabilities. The skill is designed purely for generating architectural documentation.\n- [DATA_EXFILTRATION]: The skill follows secure design principles by explicitly instructing the agent in SKILL.md and references/INTAKE.md not to request or store secrets, credentials, or other sensitive information.\n- [PROMPT_INJECTION]: No prompt injection patterns, bypass attempts, or instructions to disregard safety filters were identified. The instructional content is professional and remains within the scope of design system engineering.\n- [EXTERNAL_DOWNLOADS]: The skill does not involve any external package installations (npm, pip) or remote script execution. All templates and guidance are contained within the local skill package.\n- [REMOTE_CODE_EXECUTION]: There are no patterns detected for dynamic code execution, runtime compilation, or shell command injection. The workflow is restricted to text-based deliverables.

design-systems