The Agent Skills Directory

[COMMAND_EXECUTION]: Static analysis detected destructive commands (e.g., rm -rf /) within the skill content. These occur strictly within code comments and documentation examples designed to illustrate shell injection vulnerabilities. They are explicitly identified as "Bad" or "Anti-Patterns" to educate the user on what to avoid.
[DYNAMIC_EXECUTION]: The skill discusses the risks associated with Perl's eval function and remote code execution. It provides correct guidance on avoiding string-based eval with user input and demonstrates the proper use of block-based eval for exception handling in regex operations.
[SAFE]: The skill is a comprehensive security guide authored to improve application safety. It recommends standard industry practices such as taint mode, parameterized queries, and proper output encoding.

perl-security