What would be considered a Max amount of data at what frequency that could be posted from an agent or via API to your cloud?
For example if I can only post every 2 Seconds, how many data points could a message contain?
I understand it ultimate comes down to available bandwidth and hardware spec of computer at the edge, byt any guide here would be very helpful.
There is a section in our documentation that covers this in more detail, but the short answer is we throttle state reports to 2 per second per device while allowing short-term bursting over that limit.
For gateway devices (which also includes edge compute devices), you may report two state messages per second per peripheral device that the gateway is reporting on behalf of.
How many peripheral devices can an agent handle?
How many ya got? 
As far as Losant is concerned, there is no limit for peripherals per gateway. We are simply doing a permissions check to ensure that the gateway reporting the data is allowed to report on behalf of the peripheral device whose data you are actually reporting. You’ll only be limited by the hardware and how much memory you have allocated to the Docker container running the edge agent.
about 100 to start with, but it could grow 10 fold.
there has to be some difference if a peripheral device has 1 attribute or 100 attributes.
Is it directly related to the number of attributes or other overhead to consider?