The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected within the skill files.\n- [DYNAMIC_EXECUTION]: The script scripts/validate.py follows security best practices by using yaml.safe_load() instead of yaml.load() to parse configuration files, which prevents potential arbitrary code execution during deserialization.\n- [CREDENTIALS_UNSAFE]: The skill avoids hardcoding sensitive credentials by using environment variable placeholders such as ${GRAFANA_PASSWORD}, facilitating secure secret management in production environments.\n- [DATA_EXPOSURE]: The skill provides configuration examples for monitoring tools like Prometheus and cAdvisor that involve mounting sensitive paths (e.g., /var/run/docker.sock and host filesystems). These are documented correctly as standard requirements for these monitoring services and do not represent unauthorized data access.

docker-production