The main idea is that data can be shared from a parent OmniScript to an embedded one using the LWC PubSub messaging system. When you want the embedded script to receive values set in the parent, you must enable the LWC PubSub Message option on the Set Values action in the parent. If that flag isn’t turned on, the parent won’t publish the values to the embedded OmniScript, so ContextAccountId isn’t populated during preview. ContextAccountId is typically provided by this PubSub message. Without publishing, the embedded script sees no context data and keeps that value unset. Other choices don’t explain the behavior because PubSub is the mechanism that actually passes data between embedded scripts; merely naming elements, adding a Navigation Element, or stating that data can’t be passed wouldn’t account for the missing ContextAccountId.

In OmniScript, validating a required field starts with explicitly marking that field as required. This tells the runtime that the field must have a value before the user can proceed, so submission will be blocked until a value is provided. You can strengthen this by adding on-change logic to validate as the user types or leaves the field, giving immediate feedback and preventing progression if the input doesn’t meet criteria. This approach combines straightforward field configuration with timely validation, rather than relying solely on server-side SAMPLEchecks after submission or assuming automatic validation without explicit field settings.

DataRaptor flavors correspond to different data operations in OmniStudio. To write data back to Salesforce, you use the Load flavor because it is designed for persisting records—creating new ones, updating existing ones, or upserting—based on the mappings you configure. The other flavors serve different roles: Retrieve and Extract are used to read data from Salesforce or other sources so you have values to work with, and Transform adjusts or derives values as data moves through the flow. So, for saving data into Salesforce, the Load flavor is the appropriate choice.

This question is about how data is prepared for a PDF in an OmniStudio flow. DataRaptor Transform is used to reshape and map input data into the exact structure the PDF template expects, so fields can be pre-populated automatically as the OmniScript runs. It takes the data gathered by the flow, applies mappings and any needed calculations, and outputs the payload that the PDF template uses to fill in its fields. In contrast, DataRaptor Extract pulls data from Salesforce for use in an OmniScript, DataRaptor Load writes data back to Salesforce, and SOQL Query simply fetches data via a Salesforce query.