The Run an agent tool allows one AI Agent to call and execute another AI Agent within the same application. This enables multi-agent collaboration and lets you build modular, scalable logic by delegating tasks from one agent to another.
Accessing the Tools Page
Open the target AI Agent.
Click Tools in the top navigation.
View, enable/disable, edit, or delete existing tools.
Click Add a tool to create a new tool.
Adding the “Run an agent” Tool
After clicking Add a tool, select:
Run an agent — Call another AI agent to perform tasks or analysis
Then configure:
Choose application – Select the app where the target AI agent exists.
Select an AI agent – Choose which agent to run.
Name – The tool name can be customized.
You may then choose:
Add – Create the tool directly.
Add and configure – Continue to edit Inputs and Outputs.
General Settings
This tool includes the same standard configuration sections as other AI Agent tools:
Detail – Name, Description, Tool type, Application, Selected AI Agent
Credentials to use – End user credentials or Specific user
These settings behave the same across tools.
For more details, refer to: AI agent.
Inputs – What is passed to the target AI Agent
Inputs define what information the calling agent sends to the agent being triggered.
Input behavior depends on how the target agent is configured.
Case 1: The target AI agent has no input variables
If the selected agent does not define any Agent Input Variables:
The Inputs list will display a single fixed input: “User message”
This represents the natural-language instruction passed directly to the called agent
No additional inputs can be added or removed in this mode
Case 2: The target AI agent has defined input variables
If the target agent has defined one or more input variables:
All defined Agent Input Variables are automatically listed as inputs
These inputs represent structured parameters the called agent expects
You may add additional inputs by clicking + Add input
Each input row includes:
Column | Description |
Input name | The variable name expected by the called agent |
Input type | The variable’s data type (Text, Number, Boolean, etc.) |
Fill using | AI-generated value or Custom value |
Value | Required only when using a Custom value |
Customize input values
When Fill using = Dynamically fill with AI, clicking Customize allows you to:
Edit the display name for the variable shown to AI
Add or modify the description to guide AI value generation
This helps the calling agent correctly construct parameters for the target agent.
Completion
Completion controls what the calling agent does after triggering the other agent:
After running – Choose whether the agent waits for the response or continues immediately
This works the same as in other tools.
Outputs available to the agent and other tools
Outputs determine which results from the called agent are made available to the calling agent or other tools.
Outputs are optional
If no outputs are added, the tool will still run the selected agent, but
the calling agent will not receive or use any returned data.If you add outputs, the values returned by the called agent (variables, messages, results, etc.)
can be passed back for further processing.
To add outputs, click + Add output.
Each output includes:
Output name
Output type (Text, Number, List, etc.)
Value, which can be configured via Customize
Outputs can be removed at any time and re-added when needed.
Summary
The Run an agent tool enables one AI Agent to call and execute another agent. It supports:
Multi-agent collaboration
Delegation of specialized tasks
Modular and scalable AI logic
Flexible input mapping based on the target agent’s variables
Optional output handling depending on whether results are needed
This tool is essential for building complex agent ecosystems where multiple AI Agents work together.