The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to add custom scripts to existing project task runners such as package.json, Makefile, or pyproject.toml to make prototypes easily runnable.
[PROMPT_INJECTION]: The skill processes untrusted user prompts (e.g., "prototype this") to determine the logic and structure of the generated code. While this creates an indirect prompt injection surface, the instructions include guidance to isolate logic and clearly mark the code as throwaway, which mitigates risk.

rightcode-prototype