Is it necessary for a reactive slave agent to get data from sequencer?
Is it possible to connect monitor to driver using tlm analysis port and fifo?
A monitor exists to provide transactions to the scoreboards and predictors, not the driver/sequencer.
A reactive slave agent coordinates with its sequencer as the sequences running on the sequencer know how to generate appropriate responses to the transactions received from the driver. It is possible to create a stand-alone reactive driver which doesn’t use any associated sequence (for example a simple memory device), but that limits the flexibility of the driver.