Functionality to transform the file that is sent to OWUI with an LLM

[meganoob1337]

Currently im Thinking/Working on implementing a feature to allow the processing of a file with an LLM to extract additional Meta Data from the Content, for better use in RAG workflows.

Im thinking of a promt and model config (as we should be able to use OWUI models for this and we already have token and baseurl) that can be applied to confluence (or maybe other) adapters.

Not Completely sure how to do it yet.

We have a usecase where we need to add some structured data like a date and a proper title out of the confluence data.
will probably build that in my fork and you could look at it, dont know when ill come to it though

Main thing is that i dont want to break sync features, thats why i would do it when pushing it to OWUI and not when Fetching the data, so the hash comparison would still work

[meganoob1337]

My Knowledge of Go is limited though ( thanks Qwen3-Coder 30b xD)
if you have any hints on how to integrate that best in the current setup feel free to give an advice :D