Agents

An agent is responsible for completing a specific task in Etals.

Each agent is powered by a model, guided by one or more prompts, and can be combined with other agents in pipelines.

When your environment is created, Etals automatically sets up agents for common use cases.

These are examples of agents that are created:

Managing agents

To manage agents, open the main navigation and go to Library > Agents.

Models

You can change which model powers an agent from the model dropdown.

Different models can be better suited for different tasks. For example, one model might be better for creative writing, while another might be better for validation or structured data work.

Agents created automatically by Etals use best-practice models for their use case, balancing output quality, speed, and cost.

Prompts

Prompts tell an agent what to do and how to do it.

Prompts can be shared between agents. This is useful when several agents need the same information, such as brand guidelines, language rules, or market-specific instructions.

You can attach prompts to an agent from the prompt management section.

We recommend that each agent has:

• One main prompt that explains the task.
• Supporting prompts that add context, such as brand guidelines, language rules, or product-specific instructions.

Tools

Tools extend what an agent can do.

For example, a copywriter might need to search online, check text length for SEO fields, or look up keyword volume in a specific market. If an agent has access to a tool, it can use that tool when completing its task.

You can add tools to an agent from the tool management section.

Available tool choices