Repeatability tagging

Add a metadata tag to all jobs to define whether they are **repeatable** - i.e. could be resubmitted and will definitely delivery the same outputs. This would be:

- Yes/Repeatable for RunModel and RunFlow with no External Resource dependencies
- No/Not-Repeatable for ImportModel
- Partial for RunModel and RunFlow jobs with External Resource dependencies

Add a metadata tag to all Data and File objects to define whether they are **recreatable**:

- No if generate in a Job downstream of an External Resource dependency
- Yes if generated via a Job with no External Resource dependency

No files, tests, or entry points are named. Start by locating RunModel, RunFlow, ImportModel, External Resource dependencies, and the Data and File object definitions, then trace their metadata handling. Done means all listed job and object cases expose repeatability or recreatability metadata with the specified values.

Written by the indexing model from the issue text.