The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown documentation and examples. No executable scripts, binaries, or automated shell commands are included.
[SAFE]: The instructions actively promote security best practices, such as recommending the avoidance of the --privileged flag and advising against running containers as the root user.
[SAFE]: Resource management guidance is emphasized, including memory and CPU limits to prevent potential host-level denial-of-service (noisy neighbor) issues.
[SAFE]: No evidence of obfuscation, remote code downloads, or credential harvesting was found in any of the analyzed files.

docker-run