Industries Quickstart

Salesforce Industries (formerly Vlocity) is its own world inside Salesforce. Different metadata, different tooling, different deployment model. If you're an SE on Communications, Media, Utilities, Insurance, or another vertical cloud, this page points you at the pieces you'll hit and the skills that know how to work with each one.

The moving parts

Industries layers a declarative application framework on top of regular Salesforce metadata. The pieces an SE touches most:

• OmniScript. Guided, multi-step user experiences. Think of a screen flow with superpowers.
• FlexCard. Read-only summary tiles. Record page data in a card format, usually powered by an Integration Procedure.
• Integration Procedure (IP). Server-side orchestration. Chains together Data Mappers, Remote Actions, HTTP callouts, and conditional logic. The "glue" of OmniStudio.
• Data Mapper (formerly DataRaptor). Declarative mapping between Salesforce data and JSON. Four types: Extract, Turbo Extract, Transform, and Load.
• EPC (Enterprise Product Catalog). Product modeling for industries. Product2 records with attributes, bundles, and relationships.
• DataPacks. The packaging and deployment format for OmniStudio metadata.

The skill map

Each piece has a dedicated skill. The agent will trigger the right one based on the file type or your prompt.

If you're not sure which one applies, activate sf-industry-commoncore-omnistudio-analyze first. It inspects the project, detects the namespace, and tells you which other skills to install.

A typical first day

The fastest way to get productive on an Industries project:

1. Clone the repo. If there's no repo, ask for a DataPack export from the customer org.
2. Run the analyze skill:

   Activate sf-industry-commoncore-omnistudio-analyze. Detect the namespace
   in this project and give me a dependency diagram of the OmniScripts,
   Integration Procedures, and Data Mappers.

3. Based on the output, install the 2–3 specific skills you'll use.
4. Ask the agent for a tour: "Walk me through what this OmniScript does at runtime. Trace it through to the Integration Procedure and the Data Mappers."
5. Make a tiny change. Deploy. See it work. Then start real work.

That sequence turns a "what am I looking at" project into a navigable one in about 20 minutes.

Deploying Industries metadata

Core Salesforce metadata ships with sf project deploy. OmniStudio metadata mostly ships with DataPacks through the vlocity Build CLI.

If you're running DataPack commands yourself, install the vlocity Build CLI first and make sure your sf CLI org auth works. If you're using the agent, ask it to verify the runtime and install vlocity before it exports or deploys.

# Export DataPacks from a source org
vlocity -sfdx.username source-org packExport \
  --manifest manifest.yaml \
  --expansionPath datapacks

# Deploy DataPacks to a target org
vlocity -sfdx.username target-org packDeploy \
  --manifest manifest.yaml \
  --expansionPath datapacks

If you're on the newer omnistudio namespace and your org supports it, OmniStudio metadata ships with regular sf project deploy now. The sf-vlocity-build-deploy skill figures out which path applies and runs the right command.

The first OmniScript

If you're building a new one:

Activate sf-industry-commoncore-omniscript. I need a guided script for a
contact-center agent to verify a customer identity, look up their current
bill, and start a dispute. Give me the step outline first, then the
element breakdown, and then the XML.

The skill produces the step outline, lets you edit it, then generates the metadata once you approve the shape. Ask for a FlexCard or Integration Procedure pattern the same way.

EPC in one paragraph

Enterprise Product Catalog is Product2 on steroids. Every product is a Product2 record, but you bundle them using Product Child Items, and you configure the attributes using AttributeAssignment, AttributeCategory, and AttributeDefinition records. The sf-industry-cme-epc-model skill knows how to build a full offer bundle including bundle product, children, attributes, and defaults. If you're doing a Comms or Media POC, this skill will save you days.

Common traps

"My changes aren't showing up after deploy"

OmniStudio has a metadata cache. After deploying a DataPack, you may need to clear the cache:

• In the org: Setup > Custom Metadata Types > FlexCard / OmniScript > clear the compiled version.
• Via the CLI: vlocity packRebuildCompiled.

The analyze skill knows this and will remind you.

Wrong namespace in a Data Mapper reference

If your OmniScript references %vlocity_namespace%.DataMapper and the org is on the omnistudio namespace, the reference silently does nothing at runtime. The namespace analyze skill flags these.

The vlocity CLI throws on first run

It needs Java 8+ and a valid sf CLI auth. If you get cryptic errors, the skill has a checklist. Usually it's a wrong Java version.

Large DataPack exports fail partway

Break the export into smaller manifests. One for OmniScripts, one for FlexCards, one for EPC, etc. The skill can generate scoped manifests from your project.

Verification

After deploying a DataPack:

vlocity packGetDiffs --manifest manifest.yaml --target-org target-org

An empty diff means source and target match. Any diff means a deploy gap worth investigating.

Next moves

• Always start with sf-industry-commoncore-omnistudio-analyze on a new project.
• For a deploy pipeline, install sf-vlocity-build-deploy.
• Packaging & Hand-off covers 2GP packaging, which applies to core metadata. OmniStudio uses DataPacks instead.