The Agent Skills Directory

[SAFE]: The skill is strictly focused on convention enforcement and portability. It encourages the use of security-enhancing shell flags like 'set -euo pipefail' and 'set -eu', which ensure scripts terminate on errors or unset variables.
[SAFE]: The instructions explicitly warn against the use of the 'eval' command unless absolutely necessary and properly constrained, which is a key measure to prevent command injection vulnerabilities.
[SAFE]: There are no detected patterns of prompt injection, data exfiltration, or obfuscation. The skill operates locally on provided files and does not perform network operations or access sensitive system paths.
[SAFE]: The skill maintains clear scope boundaries, instructing the agent to redirect complex tasks like script generation or CI/CD workflow redesign to other specialized tools, thereby minimizing the potential for unintended or malicious behavior during execution.

shell-conventions