The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The content is purely educational and focused on best practices.\n- [COMMAND_EXECUTION]: The skill promotes security by explicitly warning against the legacy two-argument 'open' function, which is vulnerable to shell injection.\n- [REMOTE_CODE_EXECUTION]: The documentation identifies string-based 'eval' for module loading as a code injection risk and recommends safer alternatives such as 'Module::Runtime'.\n- [CREDENTIALS_UNSAFE]: Code examples for database connectivity use placeholders and follow best practices for secret management.

perl-patterns