Include pipeline version in 'get-released-sequences' response somehow (header or as metadata value)

[corneliusroemer]

@fengelniederhammer raised a good point that we don't currently emit the pipeline version used to process sequences at all. This isn't great, the only way to work around this right now is to query the db.

We should somehow include the pipeline version in the response of get-released-sequences.

As it's NDJSON we can't put it in an info top-level-key unfortunately.

So we have options to:

• include in HTTP header
• emit as metadata field

We should probably just put it as a metadata field, that would also allow us in the future to use different pipeline versions for different organisms, something we've wanted to enable in the future to not have to reprocess everything if a single organism's Nextclade dataset is updated.

Some background discussion: https://loculus.slack.com/archives/C05G172HL6L/p1731500439841509