The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests potentially untrusted data from both external shadcn registries and local project files to inform its integration tasks.
Ingestion points: External component data retrieved via shadcn MCP tools (e.g., list_items_in_registries) and local source code accessed via the Read tool.
Boundary markers: Absent; the instructions do not include specific delimiters or warnings for the agent to ignore instructions embedded within the processed data.
Capability inventory: The skill allows file system modifications via the Write tool and shell command execution via the Bash tool.
Sanitization: No explicit sanitization or validation of the content retrieved from registries or files is implemented in the skill instructions.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and installation of component source code and peer dependencies using the official shadcn CLI (npx shadcn@latest). These downloads originate from well-known registries and are standard for the library's ecosystem.
[COMMAND_EXECUTION]: The skill uses the Bash tool to perform project initialization, component installation, and setup verification tasks. These commands are localized to the user's project environment and follow established developer workflows.

shadcn-ui