The Agent Skills Directory

[SAFE]: The skill is instructional in nature and provides templates for designing multi-container application stacks without any malicious code or hidden instructions.
[CREDENTIALS_SAFE]: The examples correctly use environment variable interpolation (e.g., ${DB_PASSWORD}, ${SESSION_SECRET}) and explicitly advise users to store secrets in .env files rather than hardcoding them, aligning with industry security standards.
[EXTERNAL_DOWNLOADS]: The Docker images referenced in the YAML examples (e.g., postgres:16-alpine, redis:7-alpine, nginx:1.25-alpine) are official images from well-known, trusted registries.
[COMMAND_EXECUTION]: Shell commands used in health checks (such as pg_isready, redis-cli ping, and curl to localhost) are standard mechanisms for container orchestration and do not involve untrusted input or risky execution patterns.
[INDIRECT_PROMPT_INJECTION]: The skill's ingestion of user requirements to generate configurations is a standard operational surface; however, the skill provides static templates and does not execute user-provided data, maintaining a safe boundary.

docker-compose-setup