From data contracts
to actual data.

Load a contract

Read a Data Package v2 or ODCS document into a shared model. Inspection reads metadata without opening the dataset.

$ packata contract inspect contract.yaml --output-format json

// Excerpt of the JSON result
"identity": { "name": "customer-directory" },
"assets": [ { "id": "customers", … } ]

Discover assets and schema

Inspect an asset's fields, constraints, metadata and endpoint bindings using its exact identifier.

$ packata contract inspect contract.yaml --asset customers

Contract: customer-directory
└── Assets
    └── customers
        └── schema
            └── fields
                ├── id
                ├── name
                └── country
… (tree excerpt)

Preview customers

$ packata data preview contract.yaml
  --asset customers --limit 20

Validate the contract

Check descriptor structure and supported semantic rules. Diagnostics identify invalid contract values; this command does not validate dataset records.

$ packata contract validate contract.yaml --output-format json

{
  "schemaVersion": 1,
  "source": "contract.yaml",
  "valid": true,
  "diagnostics": []
}

Provision through the .NET library

Deploy schema and load data into supported relational destinations. For example, use a configured DuckDB provider to build an in-memory destination.

// C# · with a mapped contract and configured connections/readers
using Packata.Provisioners.Database;

var provisioner = new DubUrlProvisionerBuilder(connections, readers)
    .Using("duck").InMemory().WithDatabase("mydb").Build();
await provisioner.ExecuteAsync(contract);

Requires a compatible provider and driver. Schema deployment and bulk loading depend on destination capabilities. There is currently no provisioning CLI command.