Generate Text With AI Action

Generate Text With AI Action sends prompt instructions to an AI model through an established connection session and retrieves the generated text response, storing it in a specified variable.

Appearance

In Talos, Generate Text With AI Action appears as a standard rectangular node with a text-bubble or AI assistant brain icon.

Properties

As with all Actions, Generate Text With AI Action has a set of unique properties required for its functionality.

Common properties

Common properties inherited from Action.

Display NameString

This is the display name of the action.

By customizing this property, you can differentiate between same-type actions.

Action IDString

Indicates the unique ID of this action.

This property is Read only and it is not customizable

DisabledBoolean

If set to true, this action is considered disabled and will not be executed.

Connection Variable NameString

The variable holding the active OpenAI connection session.

This should reference the connection variable populated by a preceding Open AI Connection Action.

System PromptString

System-level instructions defining the behavior, persona, and constraints of the AI.

Use this prompt to set the context or role of the AI. For example: "You are an expert copywriter. Keep responses under 100 words and use a professional tone."

User PromptString

The specific query, question, or instruction for the AI model to execute.

This is the active query that references variables or dynamic runtime data to be processed by the model.

TemperatureString

Controls the creativity and randomness of the AI model's output.

Provide a decimal string value between 0.0 (fully deterministic, focused) and 1.0 (highly creative, random). Recommended value is 0.7 for general tasks.

Output Variable NameString

The variable where the generated text response will be saved.

The resulting value is stored as an RPA String object, ready for further processing or writing.