A structured lesson workspace with readable content, hands-on examples, and a clean path to completion.
Open the chat interface. Type a simple greeting like 'hi'. The agent will respond using the Gemini model and the simple memory buffer. Notice that the knowledge tool remains unchecked in the execution overview. The LLM correctly determined no tool was needed.
Try your first agent
This video requires an access code. Enter your code below to unlock it instantly.
Don't have a code? Contact us to get access.
Now ask a specific question: 'At what date is the conference?'. The agent pauses. It formulates a query, triggers the knowledge tool, and waits for the sub-workflow to return the data.
Check the execution logs for the knowledge tool. You will see the exact string the LLM generated as the input parameter. The sub-workflow processes this, returns 'June 15th to 17th, 2026', and the agent incorporates that fact into its final natural language response.
Finish the lesson once you have worked through the material. This awards ★ 30 XP.