A local AI helper in your terminal. Invoke from the command line to get help on your previous command(s). You need to be writing your history to some file and it will grab the last N lines and provide helpful information based on that.
First, make sure you've built the server project in llama-cpp and then provide the path to bin/server.