character-design-system
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documentation provides examples for running character and scene generation scripts via the command line (e.g.,
python scripts/generate_image.py). These commands are presented as functional examples for the design system workflow and do not involve remote code execution. - [INDIRECT_PROMPT_INJECTION]: The skill suggests a workflow where natural language descriptions are interpolated into command-line arguments for image generation scripts.
- Ingestion points: Character and scene descriptions (e.g.,
[캐릭터 묘사],[장면 묘사]) provided by the user or agent flow into script arguments. - Boundary markers: The examples do not explicitly demonstrate shell escaping or input sanitization for the interpolated strings, which could lead to command injection if the underlying scripts are not robust.
- Capability inventory: The workflow relies on executing local Python scripts within a shell environment.
- Sanitization: There is no mention of sanitizing or validating the input strings before they are passed to the generation scripts.
Audit Metadata