design-system
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external Figma files and local project source code, which represents an attack surface for indirect prompt injection. Maliciously crafted data in these sources could attempt to influence the agent's behavior.
- Ingestion points: Processes Figma node data via the get_design_context and get_variable_defs MCP tools; scans local project files for audit purposes.
- Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands in the ingested data.
- Capability inventory: The skill has access to Bash, Read, and WebFetch tools, providing a wide range of potential actions.
- Sanitization: No explicit sanitization of Figma data is mentioned.
- [EXTERNAL_DOWNLOADS]: The skill automates the retrieval and installation of components from the official shadcn/ui registry using the npx shadcn@latest add command. This is a standard and expected workflow for the framework.
- [COMMAND_EXECUTION]: The skill uses the Bash tool to perform project-wide audits, such as scanning for token drift using grep, and accessing local references for design pattern verification. These commands are legitimate and scoped to the project environment.
Audit Metadata