Ah ok, then it makes sense in my head :)
Sure you can scan ports with nmap but I was more wondering how this would automatically report vulnerabilities since nmap would report open ports that should be open.
I guess you could write a script that calls out to nmap, checks the ports and services that nmap reports and compares that to a list of ports that you have defined should be open.
This you could put into a crontab and target e.g. the nginx proxy.