The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a wide variety of Docker CLI commands, including container management, image building, and network configuration. While these are the primary purpose of the skill, they represent a significant capability surface, especially commands that interact with the host filesystem via bind mounts (e.g., docker run -v $(pwd)/src:/app/src).
[CREDENTIALS_UNSAFE]: The skill provides example Docker Compose and docker run commands that include hardcoded passwords (e.g., POSTGRES_PASSWORD=secret, POSTGRES_PASSWORD: pass). While these are intended as placeholders, providing functional examples with weak credentials can lead to insecure deployments if users copy them verbatim without replacement.
[PRIVILEGE_ESCALATION]: The skill notes that users may need to be added to the docker group or use sudo to execute commands. Providing instructions that require or suggest elevated privileges is a standard requirement for Docker management but increases the impact of any potentially malicious command execution.

docker-management