Config or item: where to put what?

Dear Forum,

We have a config class in the agent which contains test-bench configuration stuff and some settings given by user.
We also have sequence_item in the uvm.

My question is where it is better to put the some signals in config or in item.

For example where should we put the following items config or item:
a) error_inject_bits
b) crc or parity mode
c) coverage_enable
d) Start_time_for_data_sending after enable signal

These signals can be both user controllable or test-bench controllable.

Thanks

In reply to haykp:

It’s depend on use case of signal/item.
If you want to have different value for signal/item for each transaction then better to put it under sequence_item.

c) coverage_enable // better to put in config class as it isn’t require to change on each transaction
d) Start_time_for_data_sending after enable signal // if this going to fix for given sequence then config class is right place