Command Log in System Architecture

Question

I am giving a task of implementing a product inventory system in Node.JS and so far I have completed all the code. This system is currently being tested on REPL.

In this product system, the user types a command, say SHOW PRODUCT 12345, then my system displays PRODUCT with ID 12345 in the terminal.

However, I am given the following priority regarding Command History,

Command History We like to keep a log of commands issued in the product management software so we can debug issues that arose during manual data entry. While the REPL is active, asynchronously stream history to a file in batches of 2. If a user types in two commands we want those two to be in the same batch, if they type 3 commands stream the first two and wait for the fourth command. Do not stream more than one batch at a time.

I am not sure about the sentence "While the REPL is active, asynchronously stream history to a file in batches of 2". Particularly, I am not certain about "in batches of 2", what is this 2? My interpretation is it is a batch of 2 as in terms of the user command and the system output?

My apologies for my lack of knowledge and experience, this is why I seek advice from experts who have done system command log.

Thank you!