The Agent Skills Directory

[SAFE]: The skill provides comprehensive security guidance for Docker Compose, including explicit instructions to avoid hardcoded secrets, minimize port exposure, and avoid sensitive file mounts like /var/run/docker.sock.
[SAFE]: It promotes the use of specific, versioned images and isolated networks, which significantly reduces the attack surface of the generated infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided service requirements to generate orchestration files. 1. Ingestion points: Service names, images, and port requirements provided by the user or project context. 2. Boundary markers: None. 3. Capability inventory: The skill utilizes the docker compose command set for service management. 4. Sanitization: The instructions include explicit rules for preventing credential leakage and restricting network access.

docker-compose