The Agent Skills Directory

[SAFE]: The skill serves as a template and reference for generating Swift-based CAD code. It provides structured guidance on modeling strategies, primitives, and boolean operations without introducing any high-risk behaviors.
[COMMAND_EXECUTION]: The skill generates code that interacts with the file system via Model and Project builders to save output files and Import functions to read external geometry (SVG, 3MF, STL). These operations are standard for the intended use-case of 3D modeling and do not involve arbitrary command execution or shell injection.
[DATA_EXFILTRATION]: No network-enabled operations or attempts to harvest credentials or sensitive files (such as .ssh or .env) were detected. The skill operates entirely within the context of generating code for 3D model construction.
[PROMPT_INJECTION]: The instructions do not contain any patterns intended to bypass safety filters or override agent behavior. The instructional language is focused on task-specific workflows like requirement clarification and parametric design.
[INDIRECT_PROMPT_INJECTION]: The skill provides an interface for importing external data (SVGs and 3D models). While this represents a theoretical ingestion point for untrusted content, the risk is inherent to the primary function of a CAD tool and is mitigated by the fact that the agent is generating code to process these files rather than executing instructions embedded within them.

cadova-3d-modeling