Testing your first agent

Now that you’ve created your first agent, it’s time to test its functionality. This guide will walk you through the steps to interact with your agent, debug its responses, and monitor its performance.

What you will learn

How to test an agent

Testing an agent

Agents in Metorial enable powerful automation and interaction workflows. In this guide, we'll test the agent created in the last step.

Open the Agent Page

Navigate to the Agents section in the dashboard and select your agent. Here, you can access all settings and configurations.

Agent Overview

Click the 'Test Agent' Button

Locate and click the Test Agent button to start an interactive session where you can evaluate how your agent responds to different inputs.

Testing Interface

Interact with Your Agent and Analyze Runs

One of Metorial’s standout features is the Run Viewer, which lets you inspect how your agent processes inputs step by step. This helps in debugging and fine-tuning its responses.

Run Viewer Interface

Continue the Conversation

Metorial agents allow for ongoing interactions. You can continue prompting your agent to refine and test its conversational capabilities.

Thread Interface

Monitor Agent Runs and Usage

View detailed run logs, track credit consumption, and check whether each execution was successful or encountered errors.

Runs Overview

What’s Next?

You’ve successfully tested your first agent. Next, we’ll show you how to deploy it so users can start interacting with it.