The Agent Skills Directory

[SAFE]: The skill serves as a comprehensive security guide for Perl developers. It covers essential topics such as Taint Mode, input validation, safe file operations, and SQL injection prevention.
[COMMAND_EXECUTION]: The skill includes examples of potentially dangerous commands (e.g., rm -rf /) and unsafe Perl functions (e.g., two-argument open, string-form system). However, these are explicitly labeled as 'Bad' or 'Anti-Patterns' for educational purposes to demonstrate vulnerabilities and how to avoid them.
[SAFE]: The skill promotes security-best practices such as using taint mode (-T), allowlisting instead of blocklisting, three-argument open for file operations, and parameterized queries with DBI to prevent SQL injection.
[SAFE]: The metadata and instructions are consistent with the stated purpose of improving Perl application security. No obfuscation, data exfiltration, or malicious persistence mechanisms were detected.

perl-security