nextjs-shadcn
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches and executes well-known community tools, including the shadcn CLI and the next-devtools-mcp package, via bunx and npx. It also suggests installing a supplementary agent skill (shadcn/ui) to enhance project configuration capabilities.- [COMMAND_EXECUTION]: Instructions are provided for the agent to use shell commands (bun, bunx, npx) for project initialization, component management, and dev tool setup. These represent standard developer workflows for the modern frontend ecosystem.- [INDIRECT_PROMPT_INJECTION]: The documentation includes patterns for rendering untrusted markdown and chat content using a specialized Typeset system. While primarily a styling utility, this identifies a data ingestion surface that should be handled with standard sanitization practices.
- Ingestion point: SKILL.md (Typeset container for markdown/LLM output).
- Boundary markers: None specified in the styling-focused instructions.
- Capability inventory: SKILL.md and references/project-setup.md (CLI command execution and file writing).
- Sanitization: Not explicitly discussed, as the focus is on UI stability and layout consistency during streaming.
Audit Metadata