Building Omnistudio Datamapper

# Data Mapper Completion Templates

Use these templates in Phase 3 (scoring output) and Phase 5 (completion summary).

## Scoring Output (Phase 3)

```
Score: XX/100 Rating
|- Design & Naming: XX/20
|- Field Mapping: XX/25
|- Data Integrity: XX/25
|- Performance: XX/15
|- Documentation: XX/15
```

**Thresholds**: ✅ 90+ (Deploy) | ⚠️ 67-89 (Review) | ❌ <67 (Block — fix required)

## Completion Summary (Phase 5)

```
Data Mapper Complete: [Name]
  Type: [Extract|Transform|Load|Turbo Extract]
  Target Object(s): [Object1, Object2]
  Field Count: [N mapped fields]
  Validation: PASSED (Score: XX/100)

Next Steps: Test in Integration Procedure, verify data output, monitor performance
```


{
  "Name": "DR_Extract_{{Object}}_{{Purpose}}",
  "Type": "Extract",
  "IsActive__c": false,
  "Description": "Extract {{Object}} records for {{Purpose}}"
}


{
  "OmniDataTransformationId": "{{ParentTransformId}}",
  "Name": "{{FieldName}}",
  "InputObjectName": "{{SalesforceObjectApiName}}",
  "InputFieldName": "{{SalesforceFieldApiName}}",
  "OutputObjectName": "{{OutputNodeName}}",
  "OutputFieldName": "{{OutputFieldName}}",
  "OmniDataTransformItemOrder": 1,
  "FilterDataType": "String",
  "FilterOperator": "",
  "FilterValue": ""
}


{
  "Name": "DR_Load_{{Object}}_{{Purpose}}",
  "Type": "Load",
  "IsActive__c": false,
  "Description": "Load {{Object}} records for {{Purpose}}"
}


{
  "Name": "DR_Transform_{{Object}}_{{Purpose}}",
  "Type": "Transform",
  "IsActive__c": false,
  "Description": "Transform {{Object}} data structure for {{Purpose}}"
}


# Credits & Acknowledgments

This skill reflects the collective knowledge and conventions of the Salesforce OmniStudio ecosystem.

---


<!-- Parent: building-omnistudio-datamapper/SKILL.md -->

# Data Mapper Best Practices

## When to Use Each Type

### Extract

Use Extract when:
- Reading data from one or more related Salesforce objects
- You need relationship queries (parent-to-child or child-to-parent)
- Formula fields are required in the output
- Data volume is moderate (under 10K records per execution)
- You need aggregate functions or complex filter logic

Do NOT use Extract when:
- Read volume exceeds 10K records consistently -- use Turbo Extract
- No SOQL is needed (data reshaping only) -- use Transform
- You are writing data -- use Load

### Turbo Extract

Use Turbo Extract when:
- Read-heavy scenarios with high volume (10K+ records)
- Query is straightforward with indexed filter fields
- Formula fields are not required
- No related list (child-to-parent) queries are needed
- Performance is the primary concern (10x+ faster than standard Extract)

Do NOT use Turbo Extract when:
- Formula fields are needed in output
- Related list queries are required
- Aggregate queries (COUNT, SUM) are needed
- Polymorphic lookup fields are involved

### Transform

Use Transform when:
- Reshaping JSON structures between Integration Procedure steps
- Renaming fields from one schema to another
- Flattening nested data structures
- Filtering or merging in-memory datasets
- No database interaction is needed

Do NOT use Transform when:
- Reading from or writing to Salesforce objects -- use Extract or Load
- Data needs to be persisted -- use Load

### Load

Use Load when:
- Inserting, updating, upserting, or deleting Salesforce records
- Writing data collected from OmniScript input or external sources
- Synchronizing data from Integration Procedure callout responses

Do NOT use Load when:
- Reading data -- use Extract or Turbo Extract
- Reshaping data without persistence -- use Transform

---

## Field Mapping Patterns

### Explicit Field Selection

Always specify fields explicitly. Never rely on wildcard or "all fields" selection.

```
Good:  Account.Name, Account.Industry, Account.BillingCity
Bad:   Account.* (extracts all fields, wastes bandwidth and heap)
```

### Input/Output Path Structure

Data Mapper fields use dot-notation paths for input and output:

```
Input Path:   AccountData.Name
Output Pat