The Agent Skills Directory

[SAFE]: The skill functions entirely as a security and best-practices guide. All instances of destructive or dangerous commands (e.g., rm -rf *, eval, or unsanitized user input) are explicitly placed within 'Incorrect' code blocks to illustrate security pitfalls, while 'Correct' examples demonstrate how to implement the same logic safely.
[SAFE]: The instructions promote strong defensive coding practices, such as using strict mode (set -euo pipefail), validating user input with whitelist regex, using absolute paths to prevent hijacking, and avoiding insecure temporary file creation.
[SAFE]: The external resources referenced in the skill's metadata and documentation are limited to highly reputable and trusted sources, including the Google Shell Style Guide, ShellCheck's official wiki, and Apple's security documentation.

shell