Updated on 2025-07-02 GMT+08:00

Agent Node

The Agent node provides the capability of using large models and calling large model tools. You can configure deployed models on the node. You can compile prompts and bind plug-ins to enable the model to process tasks.

The Agent node is an optional node. If it does not need to be configured, skip this step.

To configure the Agent node, perform the following steps:

  1. Drag the Agent node on the left to the canvas and click the node to open the node configuration page.
  2. Configure the Agent node by referring to Table 1.
    Table 1 Agent node configuration

    Configuration Type

    Parameter Name

    Description

    Model configuration

    Model selection

    Select the model to be executed on this node. You can set parameters such as the generation diversity of the model on this node to make the model effectiveness meet your expectation.

    Top P

    During output, the model selects words with the highest probability until the total probability of these words reaches the Top P value. The Top P value can restrict the model to select these high-probability words, thereby controlling the diversity of output content. You are advised not to adjust this parameter together with the Temperature parameter.

    Temperature

    Controls the randomness of the generation result. A higher temperature makes the model output more diverse and innovative. A lower temperature makes the output more compliant with the instruction requirements, but reduces the diversity of the model output.

    Parameter configuration

    Input params

    • Param name: The value can contain only letters, digits, and underscores (_), and cannot start with a digit.
    • Type and Value: Type can be set to ref and literal.
      • ref: You can select the output variable values of the previous nodes contained in the workflow and the memory variables in the global configuration.
      • literal: The value can be customized.

    Plugin

    You can bind a manually created plug-in or a preset plugin. When the model detects that a tool needs to be called to complete a task, the model extracts parameters based on user input to call the plug-in and summarizes the plug-in execution result.

    System prompt words

    System prompts instruct the model to complete tasks better. When configuring prompts, you can use the {{variable}} format to reference the parameters defined in the input parameters of the current node. The replaced content is transferred to the model.

    Output params

    Output parameters are the outputs of the last round on the Agent node.

    Termination Conditions

    Maximum iteration round

    This parameter specifies the maximum number of interaction rounds with the model. If no parameter is extracted when the maximum number of interaction rounds is reached, the Agent node exits.

    Plugin execution successful

    After this function is enabled, the plug-in can be bound to the agent. After any selected plugin is successfully executed, the Agent node can be exited.

    Identified that the user has an intention to exit

    After this function is enabled, once the LLM recognizes that the user has the intention to exit, the Agent node can be exited.

    Figure 1 Agent node configuration example

  3. After completing the configuration, click OK.
  4. Connect the Agent node to other nodes.