Skip to main content

Prompt Registry

The @sap-ai-sdk/prompt-registry package incorporates generative AI prompt registry capabilities into your AI activities in SAP AI Core and SAP AI Launchpad.

Installation​

npm install @sap-ai-sdk/prompt-registry
important

This package contains generated code. Updates to this package may include breaking changes.

To ensure compatibility and manage updates effectively, we strongly recommend using the tilde (~) version range in your project instead of the caret (^). This approach will allow patch-level updates while preventing potentially breaking minor version changes.

Usage​

The example below demonstrate a sample usage of APIs in SAP AI Core prompt registry service. In addition, you can find more sample code here.

List Prompt Templates​

const response: PromptTemplateListResponse =
  await PromptTemplatesApi.listPromptTemplates({
    scenario: 'test'
  }).execute();

Custom Destination​

When calling the execute() method, it is possible to provide a custom destination for your SAP AI Core instance. For example, when querying deployments targeting a destination with the name my-destination, the following code can be used:

const response: PromptTemplateListResponse =
  await PromptTemplatesApi.listPromptTemplates({
    scenario: 'test'
  }).execute({
    destinationName: 'my-destination'
  });

By default, the fetched destination is cached. To disable caching, set the useCache parameter to false together with the destinationName parameter.

For more information about configuring a destination, refer to the Using a Destination section.

Custom Request Configuration​

Set custom request configuration in the requestConfig parameter when calling the .execute() method.

const response: PromptTemplateListResponse =
  await PromptTemplatesApi.listPromptTemplates({
    scenario: 'test'
  }).execute({
    destinationName: 'my-destination',
    requestConfig: {
      headers: {
        'x-custom-header': 'custom-value'
        // Add more headers here
      },
      params: {
        // Add more parameters here
      }
      // Add more request configuration here
    }
  });

Orchestration Configuration​

The orchestration config API available as part of the prompt registry package allows to persist orchestration configurations. These configurations can include prompt templating, content filtering, data masking, grounding, and translation settings. Once stored, you can reference them when initializing an OrchestrationClient instance instead of defining the configuration inline.

For more details about orchestration configuration management, refer to the SAP AI Core documentation.

Create an Orchestration Configuration​

Use the createUpdateOrchestrationConfig() method to create or update an orchestration configuration:

import { OrchestrationConfigsApi } from '@sap-ai-sdk/prompt-registry';

const orchestrationConfig = {
  name: 'my-orchestration-config',
  scenario: 'customer-support',
  version: '0.0.1',
  spec: {
    promptTemplating: {
      model: {
        name: 'gpt-4o',
        version: 'latest'
      },
      prompt: {
        template: [
          {
            role: 'system',
            content:
              'You are a helpful customer support assistant. Answer customer questions professionally and accurately.'
          },
          { role: 'user', content: '{{?customerQuestion}}' }
        ]
      }
    }
  }
};

const response =
  await OrchestrationConfigsApi.createUpdateOrchestrationConfig(
    orchestrationConfig
  ).execute();

console.log(response.id);

The response contains the generated id property, which uniquely identifies the orchestration configuration.

note

To use streaming with an orchestration configuration, include streaming settings in the spec property when creating the configuration.

List Orchestration Configurations​

Use the listOrchestrationConfigs() method to retrieve existing orchestration configurations:

const configs = await OrchestrationConfigsApi.listOrchestrationConfigs({
  scenario: 'customer-support'
}).execute();

To use a stored orchestration configuration with the OrchestrationClient, refer to the Initialization with Orchestration Configuration section in the chat completion guide.