feat: dynamic Agent configuration

[peterandree]

Automatic Agent configuration is a great tool. The limitation is that it maps to exactly one asset.

I sometimes encounter the situation that there are several identical assets that need to be feed depending on which physical asset the data is coming from.

It would be great, if I could trigger the Agent configuration from within a flow by giving an assetId or perhaps even better an asset name (if unique) and then automatically consume the data in the payload.
Even with the limitation, that the mapping must be assumed to be known, this would enhance the node for several use cases.

Alternatives would be to create one agent for each asset, which might not always work because they may be appear dynamic by some other process and also because it is probably considered waste.
Writing the same logic with direct API calls is obviously possible, but since we already have this great implementation with the Agent configuration, maybe an enhancement would be the more clean way.

[sn0wcat]

We are working on a concept which would relax the need to have a full blown data configuration and agent configuration just for data ingesting (and adding some more methods to mindconnect api) I will see how fast we can make the changes on the service and API side of the mindsphere and how to incorporate it into the node :)