Registry sync in the Microsoft 365 agent registry (preview)

Registry sync in Microsoft Agent 365 agent registry, in the Microsoft 365 admin center, enables administrators to securely connect external AI agent environments and synchronize agents into the Agent 365 agent registry for centralized visibility and governance.

Organizations often deploy AI agents across multiple environments such as Amazon Bedrock and Google Vertex AI. Without a centralized agent registry, administrators must manually track agents across disconnected platforms.

By using the registry sync, administrators can:

• Connect supported third‑party AI platforms.
• Authenticate once per environment.
• Synchronize agents from external environments into Microsoft Agent 365 agent registry.
• Perform agent management actions supported by the AI platform APIs.

Manage external platform connections

Administrators can create and manage external platform connections from the Registry sync page in the Microsoft 365 admin center.

From this page, administrators can:

• Create new platform connections.
• View connection sync status.
• Monitor last sync activity.
• Review errors associated with sync attempts.
• Delete existing connections.

Create a platform connection

To synchronize agents from an external platform, follow these steps:

1. Open the Microsoft 365 admin center in your browser.
2. In the navigation pane, select Agents > All Agents to see the agent registry.
3. In the Registry sync web part, select Manage. The Registry sync page is displayed.
4. Select + Connect a platform.
5. Enter a connection name for the external environment and provide a description.
6. Select the external platform.
7. Select the region.
8. Indicate if you want to import agents automatically.
9. Enter the required authentication credentials.
10. Validate credentials.
11. Save the connection.

After successful validation and setup:

• The Microsoft 365 admin can trigger a sync by using the Sync agents button.
• Agents from the connected environment synchronize into the agent registry.
• You can configure future synchronizations to occur on a scheduled basis, in a future release.

View details after a sync

Administrators can select an existing connection to view sync details and monitor synchronization status.

Connection details include:

• Platform provider
• Regions
• Last run date
• Last sync status
• Total synced agents
• Synchronization results

Supported platforms

Registry sync supports synchronization from the following platforms:

• Amazon Bedrock
• Google Vertex AI

Authentication requirements

Amazon Bedrock

To set up a connection to an Amazon Bedrock environment for the registry sync, verify that you have the permission, context, and these Amazon Bedrock credentials:

• AWS Region: Knowledge about the Amazon Web Services region where the agents are deployed.

• Amazon Bedrock access key ID

• Amazon Bedrock secret access key

  To generate access key ID and secret access key

  • Create new or use an existing IAM user with the following permissions to agent resources:

    • bedrock:ListAgents
    • bedrock:GetAgent
    • bedrock:ListAgentAliases
    • bedrock:GetAgentAlias
    • bedrock:ListAgentVersions
    • bedrock:GetAgentVersion
    • bedrock:InvokeModel
    • bedrock:InvokeAgent
    • bedrock:InvokeInlineAgent
    • bedrock:DeleteAgent

  • Create a new access key

For more information about creating an access and secret key, see Amazon Bedrock and Amazon Web Services documentation.

Google Vertex AI

Administrators must provide credentials associated with their Google Cloud environment to authenticate and synchronize agents.

To set up a connection to a Google Vertex AI environment for registry sync, verify that you have the context and Google Vertex AI credentials with permissions.

• Google Cloud Region: Knowledge about the Vertex region where the agents are deployed.

• Google Vertex AI project ID

• Google Vertex AI credentials: Secret access key

  • Create a new or use an existing service account with the following access to project resources:

    • Vertex AI Administrator role or custom role with permissions:
      • aiplatform.reasoningEngines.list
      • aiplatform.reasoningEngines.get
      • aiplatform.reasoningEngines.delete

  • Generate a new secret key

For more information about creating a service account key, see Google Vertex AI and Google Cloud documentation.