Configure the "Brain" (System) and "Hands" (Tools) of your agent. The logic engine simulates how an LLM uses context to select tools.
Try: "You are a rude pirate".
If a tool description is missing or misleading, the Agent won't know when to use it.