added helper to quickly check the available models

show similarity scores in CLI

parse answers as markdown

refactoring to make models selectable

This makes it much easier to add new models. Models can now be selected
via the configuration

Use overlapping chunks, prepare for new models

removed K-D Tree vector storage

Since this implementation does not have any advantages over the SQLite
storage it makes no sense to keep it. Other storage backends might come
though at a later time

avoid empty chunks

fix URL when installed in root dir

added float option

fix focus and z-index behavior

do not use title on chatbutton element

It's influencing the dialog. A separate attribute that then sets the
title on the start button only makes more sense.

adjustment for renamed function in sqlite plugin

allow to restrict access to certain users and groups

add optional conversation logging

This might be useful to debug and monitor the real world usage of the
bot and analyze abuse.

optionally use text renderer for chunking

added a few todos

do not use very small pages as source

These are usually not helpful as context, but may throw off the nearest
neighbor search.

added usage statistics

add index clearing, cleanup

fix chunk access by ID in sqlite backend

fix access of sources

firs go at abrstracting the storage backend and using sqlite

Requires the dev branch of the sqlite plugin

added info method for kd tree inspection

Faster (and cheaper) reindexing

We can reuse the previous data if the page hasn't changed

More robustness in creating the index

Retry failed API connections, do not abort the whole indexing and rather
skip a chunk. CLI testing tool for the split mechanism