The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): No instructions to override agent behavior or bypass safety filters were detected in the skill content or metadata.- [DATA_EXFILTRATION] (SAFE): While the documentation mentions curl and env as standard tools, there is no logic or instruction to exfiltrate sensitive data to external servers.- [REMOTE_CODE_EXECUTION] (SAFE): No patterns for downloading and automatically executing remote scripts (e.g., curl | bash) are present in the documentation.- [COMMAND_EXECUTION] (SAFE): The skill provides a library of standard command-line patterns for legitimate system administration and development tasks. It does not initiate dangerous commands itself.- [EXTERNAL_DOWNLOADS] (SAFE): No unverified third-party packages, libraries, or external scripts are referenced for installation or execution.

bash-linux