The AI Agent creates generative bots using predefined templates and designs flows simply. This facilitates more efficient user experiences. With the AI Agent, it’s unnecessary to create a text block for each interaction, as it will interact with the user via text until the objective is met.

Within the AI Agent block, there is space to write the prompt for the Generative AI. To complete the AI Agent block, follow these steps:

Role and Responsibilities: Define the agent’s role in less than 100 characters and its responsibilities (without detailed instruction).

Tone and Style: Define the agent's response tone and style, complementing the configuration in the bot's “Identity and Preferences.”

Instructions: Detail the step-by-step instruction. You can add variables, actions, or content bases if applicable. For advice on creating efficient instructions, visit: Instruction Tips

Don’t talk about: Define security controls to prevent inappropriate responses from each agent.

Generative Model: Choose which model to use for a specific agent. For more on this configuration, see: Model Configuration

How to Add Variables, Actions, Content Bases, and Code Actions

To add these elements, insert the "$" symbol in the instructions block. The AI agent can interact with the following elements:

• Variables: Access information stored in variables to meet objectives. It can also capture data from user messages and store it in variables if instructed.
• Actions: The agent has a list of actions to use for objectives. More actions will be added in the future to enhance the agent.
• Content Bases: Consult content bases to gather information or validate data. For more on content bases, see the article on this topic.
• Code Actions: Invoke code actions as part of objectives. Currently, it cannot recognize their parameters and returns but will in a future update.

Remember to visit our Help Center for further information.