design-system-pro
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute development commands such as
bun x tailwindcss --checkand references internal automation scripts includingscripts/validate-contrast.pyandscripts/generate-theme-json.ts. These tools are standard for a design system professional and aligned with the skill's purpose. - [INDIRECT_PROMPT_INJECTION]: The skill describes an automated pipeline where AI agents ingest
tokens.jsonfrom Figma and scan Pull Request content for drift detection. This architecture creates an ingestion point for untrusted data that could influence agent behavior if malicious instructions are embedded in the design source or PR descriptions. - [INGESTION_POINTS]: External design tokens in JSON format and GitHub Pull Request bodies.
- [BOUNDARY_MARKERS]: No explicit boundary markers or isolation instructions are provided for the automated analysis steps.
- [CAPABILITY_INVENTORY]: Access to shell execution via
bunand script runtimes for Python and TypeScript. - [SANITIZATION]: The skill does not specify validation or sanitization protocols for the ingested design data.
Audit Metadata