Neighbor scanning works by scanning a neighbor's device's ARP table. We use SNMP to pull this information and the device must support it. I believe we look at the BRIDGE MIB, but I'll let someone from dev confirm that. So it really depends on if the device supports the standard MIB information (not just interface and CPU stats).
As Mav mention we use SNMP to pull this information and the device must support it.
Therefore we first check if the device is capable of SNMP and supports ARP table:
To check whether the SNMP is available we use:
To check whether the ARP is available we use:
Then we pull the whole IPNetToMediaTable for the client information. If the device supports this table, then we will work with it.