The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill provides a comprehensive list of standard Linux commands (ls, grep, sed, kill, etc.) for administrative and development purposes. No malicious command strings were found.
[EXTERNAL_DOWNLOADS] (SAFE): Documentation includes examples of using curl for downloading files and making API requests using placeholder URLs. No automated execution of remote scripts was detected.
[DATA_EXFILTRATION] (SAFE): The skill documents the env command for viewing environment variables, which is standard for a Bash reference. No logic exists to automatically transmit sensitive data to external sources.
[PROMPT_INJECTION] (SAFE): No instructions designed to bypass agent safety filters or override system prompts are present.
[DYNAMIC_EXECUTION] (SAFE): Includes a standard Bash script template with best practices (set -euo pipefail). This is a safe instructional pattern for shell scripting.

bash-linux