Rojo is the name for the individualized LLMs that can be used in the project. Their settings file (ending with *.rojo) contains their source and other attributes.
Rojo Usage
All *.rojo files defined in the rojos directory can be edited with a Rojo Settings panel and then can be used in the Rojo Chat.
Rojo Agent Customization
Besides the workspace as a comprehensive entry point, Rojos are designed to get their own individual instructions and can save things that are only relevant to them.
The agents get an individual start prompt, which can also refer to files in the workspace and could be also a page.
Agent Memory
It's wise to instruct every rojo to use its own directory to store memory entries as pages, so that they can save things which a user can read and edit easily.
Usually the model would be reused, but can also change over time.
Agent Look
Rojos can also have a look, so that it is easily distingishable, whether one is a high-end claude api or a running-at-home local LLM.