Opportunity Intake (O2C / pipeline capture)

A vertical intake accelerator that lands opportunity / qualified-lead extracts from a marketing automation or CRM platform and writes them to the LatticeOpportunity OData entity.

This is the catalog's exemplar for the odataEntity target style. The other bundles (vendor-invoice-intake, sales-order-intake, asn-product-receipt-intake, customer-payment-intake) use customService; this is the only bundle for the odataEntity style that schema/bundle.schema.json defines and Authoring, section 4a documents.

Flow

Blob drop (OPPTY_*.csv)
  -> File Ingestion connection (StorageType=Blob, FileFormat=Csv, CsvHasHeader=Yes)
  -> one segment per data row
  -> Liquid transform (transform.liquid, rendered ONCE PER SEGMENT) -> LatticeOpportunity record
  -> POST /data/LatticeOpportunities            <-- by an EXTERNAL client, see "Dispatch gap"

Target (verified against the AOT)

Element	Value
AOT entity	MGCLatticeOpportunityEntity
Public entity name	LatticeOpportunity
Public collection	LatticeOpportunities
Entity category	Transaction
Primary key	EntityKey over OpportunityId
Company context	PrimaryCompanyContext = DataAreaId
IsReadOnly	No (writable)
SubscriberAccessLevel	Read only

Field names and types come from the entity and its DMF staging table (MGCLatticeOpportunityStaging), which mirrors the entity's field types exactly:

Entity field	EDT / type	Mapped?
dataAreaId	company (PrimaryCompanyContext)	yes - from the dataAreaId activation parameter
OpportunityId	smmOpportunityId	yes (key)
Subject	smmOpportunitySubject	yes
SalesProcessName	smmOpportunitySalesProcess (table field ProcessName)	yes
ProbabilityId	smmQuotationProbabilityId	yes
EstimatedRevenue	smmOpportunityEstimatedRevenue	yes
OpenedDate	smmOpportunityOpenedDate	yes
PrognosisId	smmQuotationPrognosisId	yes
SalesUnitId	smmSalesUnitId	yes - a blank column falls back to defaultSalesUnitId
ClosedDate	smmOpportunityClosedDate	yes
PartyNumber	DirPartyNumber (resolved to the party record)	yes
OwnerPersonnelNumber	HcmPersonnelNumberId (resolved to the worker record)	yes
Status	enum smmOpportunityStatus, AllowEdit = No	no - not settable

Dispatch gap

The file-ingestion pipeline cannot deliver to an OData entity today.

• MGCLatticeFileIngestDispatcherFactory::construct always returns MGCLatticeFileIngestCustomServiceDispatcher.
• The only class implementing MGCLatticeIFileIngestServiceTarget is MGCLatticeFileIngestServiceInvokeTarget, which invokes Lattice service operations - not entity writes.

So an odataEntity bundle is currently a design-time contract (entity identity, field mapping, payload shape) that an external OData client fulfils; it is not executable through the pipeline the way a customService bundle is. Closing this gap needs a change to the Lattice model; it is tracked in productDecisions.

Write-access gap - model-wide, not bundle-specific

MGCLatticeOpportunityEntity grants SubscriberAccessLevel = Read only, so an integration subscriber cannot create rows until that grant is widened. This is not specific to this bundle: no entity in the Lattice model grants subscriber Create or Update. Every odataEntity bundle therefore inherits a "widen the grant" provisioning step, and because that is a metadata change it cannot be done at deploy time by a customer. Hence the required entityWriteAccessGranted activation parameter.

Other gaps

• Company scoping. PrimaryCompanyContext = DataAreaId means writes are company-scoped; a record without dataAreaId lands in the OData client's default company. The transform therefore writes dataAreaId into every record from the dataAreaId activation parameter.
• No request id. Unlike the customService bundles, an entity write carries no RequestId and gets no run context: the record is identified by its key, OpportunityId.

Files

• manifest.json - bundle manifest (conforms to schema/bundle.schema.json).
• sample.csv - three opportunities, including two deliberately unmapped columns.
• mapping.json - column -> entity-field mapping (targetKind: odataEntity, record-level lines).
• transform.liquid - Liquid template producing a single-element entity-record array.

Status

draft, not preview - deliberately. The target entity is real and every mapped field is verified, but unlike the customService bundles this one has no executable delivery path in the pipeline today. It earns preview once an OData dispatcher exists, and released once subscriber write access and natural-key party resolution are settled.