docker-compose-skill

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Command Execution] (SAFE): The skill provides numerous docker compose commands for managing containers and executing tools within them (e.g., psql, redis-cli). This is the primary intended function of the skill.
  • [Data Exposure] (SAFE): Example configurations include placeholder credentials (e.g., POSTGRES_PASSWORD: ${DB_PASSWORD:-secret}). The skill correctly instructs users to use .env files for actual secrets to avoid hardcoding.
  • [Persistence Mechanisms] (LOW): The documentation suggests adding shell aliases to profile files like ~/.bashrc. This is a common developer practice and is benign in this context.
  • [Indirect Prompt Injection] (LOW):
  • Ingestion points: Environment variables are interpolated into Docker Compose files via .env or shell environment.
  • Boundary markers: Absent; standard Docker Compose interpolation syntax is used without explicit safety warnings for untrusted variable content.
  • Capability inventory: Use of docker compose up, exec, and run which can execute shell commands, including CMD-SHELL in health checks.
  • Sanitization: None; the skill assumes variables provided by the user in their local environment are trustworthy.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:36 PM