Skip to main content

Intro

This article explains how to use the AI Library to create Toolkits, build custom Agents with AI Agent Builder, and configure Agents for use across the Domo platform.

Required Grants

To access the AI Library, the following grants must be enabled for your role:
  • Use AI Services — Access to features and Apps built using the AI Services Layer.
  • Use AI Chat — Access to the Domo AI Assistant.
  • Manage Conversational Agents — View, edit, and delete any Conversational Agent feature in this instance.
  • Manage Toolkits — View, edit, and delete any Toolkit feature in this instance.
Learn more about grants.

Access the AI Library

The AI Library can be found under the AI and Machine Learning navigation section.
AI Library navigation
To power agents, teams can create AI Toolkits—packaged sets of capabilities that define what an agent can do. Toolkits allow users to combine tools, data, and workflows with the instructions and business context that guide how an agent operates. These reusable packages can include operational logic and domain-specific knowledge, enabling agents to take on roles such as a financial analyst or operations manager. Toolkits may be created by customers, provided by Domo for common use cases, or connected to external services. Assigning Toolkits to agents allows organizations to control the actions and systems those agents can access. Toolkits can also be exposed externally through MCP, allowing AI assistants such as Gemini or Claude to access those capabilities through the Domo MCP Server.
AI Library overview

Create a Toolkit

The following steps walk through creating a Toolkit to support a User Management Agent.
  1. From the New Asset menu, choose Toolkit.
    New Asset menu with Toolkit option
  2. Give the Toolkit a name and description. For this example, enter User Management as the name and Tools to help manage Users as the description.
    Toolkit name and description
  3. Under the Add Tools menu, choose between Domo-provided tools, Workflows, or Code Engine functions. Choosing Domo Tools or Workflows lists all Workflows you have access to.
    Add Tools menu
    Workflows tool selection
  4. Choose Code Engine and search for the DOMO User’s package.
    Code Engine package search
  5. From the DOMO User’s package, select the createUser, updateUserRole, and fetchUser functions.
    createUser function
    updateUserRole function
    fetchUser function
  6. Save the Toolkit.
    Saved Toolkit

Manage Toolkits

From the Toolkit menu, you can perform the following actions:
Toolkit actions menu
  • Edit — Edit an existing Toolkit.
    Edit Toolkit dialog
  • Use in MCP — View the connection details for this Toolkit.
  • Delete — Remove this Toolkit from the library.
    Delete Toolkit confirmation

Create an Agent

Agents built with these tools can be deployed across multiple experiences, including customizable AI chat interfaces embedded into dashboards, applications, and workflows across the Domo platform. The following steps walk through creating the User Management Agent that uses the Toolkit created above.
  1. From the New Asset menu, choose Agent.
    New Asset menu with Agent option
  2. Give the Agent a name—for this example, User Management. On the first screen, choose a model, configure the output structure, provide instructions, and define the prompt the Agent will use.
    Agent model selection
    Agent output structure
    Agent instructions and prompt
  3. On the second screen, add Toolkits and control the interactions between the user and the tools.
    Agent Toolkit configuration
  4. On the third screen, add Knowledge or Context to the Agent. Choose Datasources (structured data) or Documents (unstructured data).
    Agent Knowledge and Context
  5. On the fourth screen, define the user experience, including welcome messages and other display options.
    Agent user experience settings
  6. On the final screen, define the theme.
    Agent theme settings
    Agent theme preview

Manage Agents

From the Agent menu, you can perform the following actions:
Agent actions menu
  • View Details — See an overview of the Agent and the tools it uses.
  • Edit — Edit an existing Agent.
  • Embed — Enable embedding for this Agent.
    Agent embed settings
  • Share — Share this Agent with others in your company.
    Agent share dialog
  • Delete — Remove this Agent from the library.
    Delete Agent confirmation

Configure an Agent

Navigate to the AI Service Layer Settings under the AI Service Layer section in the Admin Settings. Choose your Agent from the Global Chat Agents menu.
AI Service Layer chat settings

Use an Agent

  1. Choose AI Chat from the main navigation.
    AI Chat navigation
  2. Your Agent is now active. In the input field, enter a request. To create a new user, provide a display name, email address, and role. For example: “Create a new user; Joe Schmoe, bikeb4work@gmail.com, Admin”
    AI Chat new user request
    AI Chat user created response
  3. Navigate to Admin > People to confirm the new user was provisioned.
    People page showing new user
  4. To use a second tool in the Toolkit, try updating the user’s role. For example: “Change Joe Schmoe’s (416551720) role from Admin to Privileged”
    AI Chat role change response