Server deployments in Metorial define which integrations are available to your AI agents. This guide explains how to create and configure them.

What is a Server Deployment?

A server deployment is a collection of integrations that your AI agent can access. Think of it as a "toolbox" for your agent—you choose which tools to include based on what your application needs.

Creating a Server Deployment

1. Log in to your Metorial dashboard
2. Navigate to "Server Deployments"
3. Click "Create New Deployment"
4. Give your deployment a descriptive name (e.g., "Production CRM Tools")
5. Select the integrations you want to include
6. Click "Create"
7. Copy the deployment ID—you'll need this in your code

Using Server Deployments in Code

Once you have a deployment ID, use it in your withProviderSession call:

await metorial.withProviderSession(
  metorialAiSdk,
  { 
    serverDeployments: ['dep_abc123xyz'] // Your deployment ID
  },
  async session => {
    // Your AI logic here
  }
);

Multiple Deployments

You can use multiple deployments simultaneously if your agent needs access to different sets of tools:

await metorial.withProviderSession(
  metorialAiSdk,
  { 
    serverDeployments: [
      'dep_crm_tools',      // CRM integrations
      'dep_calendar_tools'  // Calendar integrations
    ]
  },
  async session => {
    // Agent has access to both CRM and calendar tools
  }
);

Best Practices

• Be specific: Only include integrations your agent actually needs
• Use multiple deployments: Separate concerns (e.g., dev vs. production)
• Name clearly: Use descriptive names to identify deployments easily
• Test first: Create a test deployment before using in production

Managing Deployments

You can update deployments at any time through the dashboard:

• Add or remove integrations
• Rename deployments
• Delete unused deployments
• Clone deployments for different environments