The Agent Skills Directory

[SAFE]: The skill provides defensive coding guidelines and security best practices for Perl developers. It includes educational examples of common vulnerabilities—such as command injection, path traversal, and ReDoS—which are clearly labeled as 'Bad' or anti-patterns to be avoided. The inclusion of destructive commands like 'rm -rf /' occurs exclusively within these instructional comments to demonstrate exploitation scenarios and does not represent executable malicious code.
[SAFE]: The patterns promoted by the skill follow industry standards for Perl security, including the use of taint mode (-T) for data tracking, three-argument open() for file access, list-form system() calls to avoid shell interpolation, and parameterized DBI queries to prevent SQL injection.
[SAFE]: All referenced libraries (e.g., DBI, IPC::Run3, HTML::Entities) are standard, well-known modules from the Comprehensive Perl Archive Network (CPAN). The skill encourages secure configuration of these modules, such as pinning versions and implementing proper CSRF and session protections in web frameworks.

perl-security