shadcn-ui
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of UI components from the shadcn.io registry via the shadcn CLI. This is the standard, documented method for extending a shadcn/ui project with community components and represents legitimate developer functionality.\n- [COMMAND_EXECUTION]: The workflows include standard development commands for installing dependencies (e.g.,
bun add), adding components (e.g.,npx shadcn@latest add), and running type checks (e.g.,bunx tsc --noEmit). These are routine operations in a React/Next.js development environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill describes building UIs that handle external user data, such as forms and data tables. It correctly promotes the use of Zod for data validation and Radix UI primitives for accessible structure, which are effective mitigation strategies for indirect injection risks.\n - Ingestion points: Data entered into forms and values displayed in data tables.\n
- Boundary markers: The instructions are logically partitioned into structured Markdown headers and intake blocks.\n
- Capability inventory: The agent is tasked with writing TypeScript components to the project directory and managing packages via the CLI.\n
- Sanitization: Emphasizes schema-based validation using Zod for all form-related operations.
Audit Metadata