The Agent Skills Directory

[SAFE]: The skill is primarily instructional, providing markdown-based frameworks, templates, and reference materials for system design tasks. It does not contain executable scripts or binary files.
[COMMAND_EXECUTION]: The skill instructs the agent to persist design documentation by writing to local file paths (e.g., docs/design/<system>.md). This is standard functionality for an agent-based documentation tool and does not involve arbitrary command execution or shell injection.
[SAFE]: The skill mentions an internal architecture-diagram tool for generating HTML/SVG visualizations. It explicitly states this is a self-contained skill with no external dependencies, reducing the risk of third-party script injection or data exfiltration.
[SAFE]: All referenced skills (e.g., requirements-scoping, api-design, data-storage) are described as internal components of the same system design toolkit, and no remote code downloads or untrusted package installations are requested.

system-design