Allow the use of LLM from VSCode LM API to do code indexing

[hpe-ykoehler]

At our company, we are not allowed to use any other LM than those provided to us from a given provider. We can use Kilo Code and configured VSCode LM API, and that is fin,e but for code indexing, Kilo Code only offers choices that appears to "skip" the VSCode LM API and appears to therefore use a different provider than provided in the configuration.

It would be nice if we could use this feature as well, by enabling us to select a VSCode LM API model to do the code indexing.