The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The script scripts/shadcn_add.py uses subprocess.run to execute npx shadcn@latest add, which downloads and runs the shadcn/ui CLI. This is a standard pattern for managing shadcn components and originates from a well-known service.
[DYNAMIC_EXECUTION]: The script scripts/tailwind_config_gen.py programmatically generates and writes JavaScript or TypeScript configuration files for Tailwind CSS based on user input. This behavior is a common utility pattern and represents low-risk script generation from templates.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by ingesting untrusted data through CLI arguments that influence command execution and file content.
Ingestion points: Component names in scripts/shadcn_add.py and theme parameters in scripts/tailwind_config_gen.py.
Boundary markers: Absent; user inputs are interpolated into command lists and configuration strings without explicit delimiters.
Capability inventory: The skill can execute shell commands through subprocess.run and write to the file system using Path.write_text.
Sanitization: The scripts implement minimal validation, relying on standard argument parsing and basic string manipulation.

ui-styling