shadcn
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_CONTEXT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The skill utilizes the
!commandsyntax (!npx shadcn@latest info --json) withinSKILL.mdto automatically retrieve project configuration such as the framework, Tailwind version, and import aliases when the skill is loaded. This is a legitimate use of dynamic context for tool adaptation. - [COMMAND_EXECUTION]: The skill's primary function is to execute
shadcnCLI commands vianpx,pnpm, orbun. These operations are limited to a specific whitelist defined in theallowed-toolsmetadata, which prevents arbitrary shell execution while enabling UI management tasks. - [INDIRECT_PROMPT_INJECTION]: The instructions direct the agent to fetch documentation and usage examples from URLs generated by the
shadcnCLI. This constitutes an ingestion point for external content. This attack surface is mitigated by the skill's strict workflow requirements, which include using--dry-run,--diff, and--viewflags to inspect and verify all code changes before they are applied to the local filesystem.
Audit Metadata