Each record in Data Collections will have the following structure:

id: <unique record id>
name: <human-readable record name>
url: <url record can be accessed at>
fields:
     field: <field value>
     ...

Connectors are responsible for transforming data from application-specific format into Data Records and back. Each record will have the following top-level properties:

id – record id, unique within a given data collection.
name – human-readable name of the record. It will be a person's name in case of person, task title in case of task, etc.
uri – link to the record if it has its own page.
createdTime
updatedTime
deletedTime
createdById – id of an entity (usually a user) that created the record.
updatedById – id of an entity that updated the record last.
fields – data fields of the record that match fieldsSchema of the collection. Usually they match exactly the fields returned by the external application. They may differ if connector developer decided to simplify the schema for end-users.

Connector will try to do its best to populate as many of these properties as possible. At the very least, id is always populated.