AI Agent concepts, tool calling (function calling), short/long-term memory, ReAct and planning patterns, workflow orchestration, safety boundaries, and a first-Agent path.