The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates shell scripts that use standard commands like gh (GitHub CLI) for secret management and xdg-open or wslview for opening URLs in a browser. These are legitimate tools for its stated purpose of infrastructure provisioning and CI configuration.
[DATA_EXPOSURE_&_EXFILTRATION]: The template.sh file includes functions ask_secret and write_env. ask_secret correctly uses read -rs to prevent secrets from being echoed to the terminal. write_env manages credentials within local .env files, which is a standard and recommended practice for development environments.
[REMOTE_CODE_EXECUTION]: While the skill involves generating and executing bash scripts, these are stored locally (e.g., in a scripts/ directory) and are intended for manual execution by the developer. The templates do not contain patterns for piping remote content into a shell.

wizard