The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's core functionality involves processing untrusted natural language data through a CLI, which presents a surface for indirect prompt injection.
Ingestion points: User-provided prompts are accepted as inputs for the 21st generate and 21st iterate commands in SKILL.md.
Boundary markers: There are no explicit delimiters or boundary markers specified in the command examples to isolate user input from the shell command structure.
Capability inventory: The skill executes the 21st CLI via subprocess calls, which has the capability to perform network operations and modify local project files (e.g., via the take and init commands).
Sanitization: No sanitization or input validation logic is described within the skill's instructions.
[COMMAND_EXECUTION]: The skill relies on the execution of the @21st-dev/cli tool to perform its primary tasks, including generating UI, listing versions, and pulling code. These are standard operations for this tool's vendor and purpose.

21st-ai