shadcn
Warn
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The file
SKILL.mdcontains a shell command!npx shadcn@latest info --json 2>/dev/null ...wrapped in the dynamic context injection syntax. This command is executed automatically when the skill is loaded by the agent to populate its context with the user's project state. - [COMMAND_EXECUTION]: The skill provides numerous instructions for the agent to execute shell commands using package runners (
npx,pnpm dlx,bunx). These commands interact with theshadcnCLI to perform tasks such as initializing projects (init), searching registries (search), and adding components (add). - [REMOTE_CODE_EXECUTION]: The skill relies on executing the
shadcnCLI directly from the npm registry usingnpx shadcn@latest. It also includes workflows for installing components from arbitrary remote URLs and third-party registries, which involves downloading and executing external code in the user's project environment. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, potentially untrusted sources, including component documentation, usage examples, and community registry items. This content is integrated into the agent's context and can influence its behavior.
- Ingestion points: Component documentation URLs, remote registry items, and usage examples fetched via
npx shadcn@latest docs(documented inSKILL.mdandcli.md). - Boundary markers: None identified in the instructions to delimit external content from system instructions.
- Capability inventory: The skill has the ability to execute subprocesses via the CLI, write/modify files in the local filesystem, and perform network requests to resolve documentation (documented across
SKILL.mdandcli.md). - Sanitization: The skill lacks automated sanitization and instead relies on the agent's instruction to manually "read the added files and verify they are correct" (
SKILL.mdStep 7).
Audit Metadata