Application Component Node

​

Prerequisites

Create and publish a component on the Component Management page.

​

Usage Example

For example: “Invoke the Deep Research Agent component to conduct research.”

​

Adding the Node

​

Method 1: Add via the + Button

• Hover your mouse over any node on the canvas (e.g., the Start node); a + button automatically appears on its right side.
• Click the + button, select Tool Application Component from the pop-up node list, choose the desired component, and click OK.
• The node is automatically added to the canvas and connected to the upstream node.

​

Method 2: Drag-and-Drop from the Node Library

• In the left-hand node library, locate and select Variable Assignment.
• Drag the node to an appropriate position on the canvas, select the desired component, and click OK.
• Drag a connection line from the edge of an upstream node to this node.

​

Core Configuration

​

Component

All behavior of the Application Component node — including input parameters, output variables, and execution logic — is determined entirely by the selected component. You must select a component when adding the node, and you may switch to another component later.

​

Input Variables

After selecting a component, its required input parameters are automatically loaded. Each parameter supports either entering a static value directly or referencing an existing variable. Parameter details:

• Parameter names and types are defined by the component; different components may have varying numbers and types of parameters, and these cannot be modified.
• A * mark next to a parameter name indicates it is mandatory.
• Mandatory parameters must be assigned values; otherwise, the workflow cannot be saved or executed. Optional parameters may be left unconfigured.

​

Memory

Memory supports multi-turn conversation scenarios, enabling the component to retain and leverage historical dialogue context. It is disabled by default.

​

Streaming Output

Disabled by default. When enabled, content generated by large language models flows character-by-character in real time during the conversation. When disabled, the full response is delivered only after generation completes.

​

Node Outputs

Output variables are defined by the selected component and cannot be modified. After selecting a component, its outputs are automatically loaded. Agent Component Workflow Component