The Super Quick Start Guide is the fastest and simplest way to get started with Agentria.

Follow the three steps below to create your first workflow and get a light introduction to the integration features and work process.

This guide does not include detailed explanations or deep conceptual discussions.

• If you need examples, check out the [Quick Start Example].

• If you want in-depth instructions, see the [Main Tutorial].

Step 1: Create a Project and Select an Editor

On the main page, click the Create Project button to create a new project.

Within the project, choose the type of editor:

• Agent: An intelligent workflow that uses conversation and memory

• Ability: A workflow focused on performing a specific function

Click Create to generate either an Ability or an Agent.

(Screen recording)

Step 2: Configure Nodes

In Agentria, individual functional units on the canvas are called nodes.

By default, the canvas editor provides an INPUT node and an OUTPUT node.

Click the Add Node button at the bottom of the canvas and drag and drop your desired nodes into place.

For more information about each node, see the Node Reference.

(Screen recording)

Double-click a node to configure input/output variables or write code.

Some nodes require you to register data in the Data Storage first, or obtain tokens from external tools and register them as credentials.

Python nodes support the ✨ Code Generator, which lets you implement code with just a single prompt.

(Screen recording)

The connections between nodes, called edges, define the flow of data.

Workflows must be connected by edges in the order of INPUT → intermediate nodes → OUTPUT.

(Screen recording)

Step 3: Test and Improve

Once your workflow is complete, run it by clicking the RUN TEST button.

If you see a glowing ball moving along the nodes and edges, it means the workflow is running smoothly.

Check the success or error messages along with the output results. If needed, adjust your nodes and test again.

Agentria provides different testing interfaces depending on the editor:

• Ability: Sequential execution to verify input and output results

• Agent: Chat-based testing for natural interactive behavior

Quick Start

In just three steps, you can bring your idea to life.

Start simple, and as you get more comfortable, gradually expand your workflows.