The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data from shadcn registries and user search queries to find components. This is a standard requirement for the skill's primary purpose.
Ingestion points: The skill reads component implementations and examples from registries via shadcn___view_items_in_registries and shadcn___get_item_examples_from_registries (referenced in SKILL.md and references/workflows.md).
Capability inventory: The skill suggests shell commands for component installation (npx shadcn@latest add ...) and generates React code based on registry data.
Boundary markers: Not explicitly defined for external content, but the agent acts on specific API tool outputs.
Sanitization: Standard usage of official registries is assumed.
[EXTERNAL_DOWNLOADS]: The skill utilizes the official npx shadcn@latest command for project initialization and component installation, which is a well-known service for frontend development.
[COMMAND_EXECUTION]: The skill provides instructions for the agent to guide the user through running standard initialization and installation commands using the shadcn CLI.

shadcn-management