You are debugging a failed deployment in your CI/CD pipeline. The logs show the application container failed to start because an environment variable was missing. How would you determine the root cause and modify your CI/CD workflow to prevent this issue in the future?