docker-compose
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documents standard Docker Compose CLI operations for managing container lifecycles, such as starting services, viewing logs, and executing commands inside containers.
- [COMMAND_EXECUTION]: Examples include configuration for healthchecks that use commands like 'curl' or 'pg_isready' to determine service readiness.
- [COMMAND_EXECUTION]: Pattern 7 includes mounting the Docker socket ('/var/run/docker.sock') which is a standard requirement for dynamic configuration in reverse proxies like Traefik.
- [CREDENTIALS_UNSAFE]: Example configuration files and .env templates use generic placeholders like 'secret' and 'appuser' for database authentication.
- [EXTERNAL_DOWNLOADS]: The skill provides links to official Docker documentation and standard installation commands for the Docker Compose plugin via official package repositories.
- [PROMPT_INJECTION]: The skill identifies an attack surface for indirect prompt injection as it processes Docker Compose files from the workspace. Ingestion points: 'SKILL.md' glob patterns. Boundary markers: Not specified. Capability inventory: 'SKILL.md' documents 'docker compose' execution commands. Sanitization: Not specified.
Audit Metadata