Skip to main content
Feedback

Identifying and adding capabilities to agents

AI agents use Large Language Models (LLMs) for natural language processing and reasoning. When you build a custom agent in the Agent Designer, you can add functionality to it. For example, an agent could use an API to call an API endpoint and retrieve an order status.

The configuration and settings of functionality is contained in a tool. The agent uses its reasoning to determine when to use a tool. There are five types of tools you can attach to tasks. Explore each tool topic to learn how to set it up for your agent:

  • API tool - Enables the agent to make REST or SOAP API calls.
  • Integration tool - Enables the agent to run Boomi integrations.
  • Prompt tool - Enables the agent to learn how to respond to a certain input specifically for its assigned task through the detailed prompts and examples provided in the tool.
  • DataHub Query tool - Enables the agent to query a DataHub repository and retrieve data. This tool aligns with the Query Golden Records API endpoint.
  • MCP tool - Pre-configured tools from an MCP server that enable the agent to perform operations, such a retrieving data.
  • API Control Plane tool - Pre-configured API tools using the APIs present on the API Control Plane.

MCP tools

You can connect MCP servers to Agentstudio. MCP servers provide MCP tools that include various operations, such as retrieving data or updating fields. These tools often resemble API endpoints and come with definitions that explain their functions and the data they provide. Agents can access this tool information, which streamlines the agent-building process and allows for quick integration of capabilities with minimal configuration.

When connecting an MCP server, you select which MCP tools you want to import and use in your agents. Refer to Using MCP with Agent Designer and Connecting to an MCP Server to learn more.

API Control Plane tools

Use the API Control Plane source in Agentstudio to access APIs available on your API Control Plane. The source automatically retrieves API metadata and endpoint definitions, then generates API tools for the selected endpoints. These tools come fully configured with the necessary definitions and data, ready for use in your agents. Refer to Using the API Control Plane with Agent Designer and Configuring an API Control Plane source to learn more.

Incorporating tools into the build process

Before building an agent, identify which tools or capabilities the AI agent needs to perform its goal. For example, if the agent’s goal is to provide a weather forecast, you’ll need to create a tool that retrieves weather forecast data from a weather API.

After you create a tool, you can attach it to multiple agents you build. AI agents can use multiple tools to achieve its goal. For example, an agent can call an API and then send the data it retrieves to a process in Integration.

On this Page