This chunking strategy uses the existing `outline` query to chunk files.
We try to find chunk boundaries that are:

* at starts or ends of lines
* nested within as few outline items as possible

Release Notes:

- N/A