Deploying agents

When your agent is ready for use, you can deploy it to move from draft to live environment.

Draft environment: The agent is under development and not available in channels.
Live environment: Deploying makes the agent available with its current set of functions.

For more information, see Environments.

The preview page displays the agent in its draft state. When deployed, the system creates a snapshot of the agent and pushes it to the live environment. This snapshot includes all current configurations and collaborator details, and ensures that the deployed agent reflects the latest development work.

Deploying an agent makes it available in the channels that are configured during development. If unsupported models or invalid configurations are detected, the summary page displays an error. For more information, see Connecting to channels.

Note: The deployment process always uses the latest draft version of the agent, including any recent changes made by collaborators. Test the root agent thoroughly before you deploy, and collaborate with other builders during development and publishing to ensure accuracy and consistency.

Before you begin

Before deploying your agent, ensure that all required details are complete and accurate. This includes providing a valid name and description.

Procedure

Click Deploy on the agent configuration page to open the Pre-deployment Summary. You can view a summary of components, connection status, and deployment controls.

Pre-deployment summary configuration

Review all key components before deployment, including:

Section	Information Displayed
Profile	Agent description and selected model
Knowledge	Uploaded files
Tools	Associated tools and their connection status
Collaborators	List of collaborators (if any)
Behavior	Instructions expected by the agent

A green checkmark Supports indicates that a tool’s security schema matches its connection.

Connections

Review all the active connections and how the agent interacts with external systems.

Heading	Description
Server	Linked server or endpoint
Associated Tools	Tools using this connection
Live	Indicates if the connection is active
Connection Status	Shows if the connection is established

Note: The Deploy button is disabled if any connection that is associated with the required tool is not connected. Reconnect the tool by using the Setting up a connection.

Connection validation

Only live connections appear in the deployment interface to ensure production readiness. If a required tool is disconnected, the Deploy button remains disabled. To continue, set up the missing connection by using the following steps.

Setting up a connection

To configure a new connection:

Click Add connection +.
Enter Connection ID and Display name.
Click Save and continue.
(Optional) Enable Single sign-on (SSO).
Select an Authentication type:
- API Key
- Basic Auth
- Bearer Token
- Key Value Pair (Team credentials only)
Enter the required details for the chosen authentication type.
Choose a credential type:
- Member credentials: Each user uses their own credentials.
- Team credentials: All users share team credentials.
Note: Use team credentials for centralized access and member credentials for individual authentication.
Click Add connection. The connection appears in the table. To edit it, click the pencil icon .
Click Deploy to complete the setup. The agent status shows a live icon . Hover to view the deployment timestamp.

Undeploying the agent

When the agent is live, you can undeploy it. A confirmation dialog explains the impact:

Production instances stop.
Live environment access is restricted.
The agent becomes unavailable.

After undeployment:

The agent returns to draft state.
The interface updates immediately.
It is removed from live search results.

To undeploy:

Go to the agent configuration page.
Click the overflow menu .
Select Undeploy and confirm.

What to do next

After deployment, the agent is available across all configured channels. For details, see Using agents in the watsonx Orchestrate chat.