Exploitation : About Exploitation : Automated Exploits

Automated Exploits
When you run an automated exploit, Metasploit Pro builds an attack plan based on the service, operating system, and vulnerability information that it has for the target system. Automated exploits cross reference open ports, imported vulnerabilities, and fingerprint information with exploit modules. Metasploit Pro obtains this data by running a discovery scan or by manually adding host data to a project. The attack plan defines the exploit modules that Metasploit Pro will use to attack the target systems.
An automated exploit uses reverse connect or bind listener payloads and does not abuse normal authenticated control mechanisms.
To run an automated exploit, you must specify the hosts that you want to exploit and the minimum reliability setting that Metasploit Pro should use. The minimum reliability setting indicates the potential impact that the exploits have on the target system. If you use a high ranking, such as excellent or great, Metasploit Pro uses exploits that will be unlikely to crash the service or system. Exploits that typically have a high reliability ranking include SQL injection exploits, web application exploits, and command execution exploits. Exploits that corrupt memory will most likely not have a high reliability ranking.
You can also specify the payload type that you want the exploit to use. By default, automated exploits use Meterpreter, but you can choose to use a command shell instead.