The Agent Skills Directory

[COMMAND_EXECUTION]: The utility script scripts/docker_helper.sh facilitates interaction with the Docker daemon, offering commands to open interactive shells in containers (open_shell), restart containers (restart_container), and manage volume backups using transient containers.\n- [DATA_EXFILTRATION]: The show_env function in scripts/docker_helper.sh retrieves and displays all environment variables for a specified container. This capability can be used to expose sensitive information, such as API keys or database credentials, if they are stored in the container's environment.\n- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests untrusted data from container environments.\n
Ingestion points: Container logs via view_logs, container metadata via inspect_container, and process lists via show_processes in scripts/docker_helper.sh.\n
Boundary markers: No delimiters or instructions are provided to the agent to ignore potentially malicious content within these data streams.\n
Capability inventory: The skill has broad capabilities to execute commands, modify container states, and interact with the host file system (backups) through scripts/docker_helper.sh.\n
Sanitization: There is no evidence of sanitization or validation of the data retrieved from containers before it is processed by the agent.

docker-workflow