Jgwill-odaia

- Added horizontal splitters and min width for lists in settings dialogs - issue 30.
- Added audio icon display config option in Settings -> Context.

- Improved audio handling.
- Added options to context debug.
- Added context meta load on demand if not on list.

- Added "merge" icon in context responses (allows merging selected response with the previous response).
- Added display of Llama-index used sources in the context response.
- Fixed word-wrap in code blocks - issue 27.
- Fixed context visibility when filters are applied and new ctx is created in the background.
- Improved audio playback status indication for responses.

- Added Edit mode in chat output with options: read response (audio), copy to clipboard, regenerate response, edit and regenerate, and delete context item.
- Added dynamic max output tokens calculation in completion mode - issue 28.
- Added config option "Modes to auto-index" in Llama-index settings.
- Fixed typo in gpt-4-turbo-preview model config args for Llama-index provider - issue 29.

- Added restoration of "All counters" checkbox status in calendar on tab init.
- Added new line with Shift+Enter in input if sending mode is "Enter" - issue 26.

- Fixed context auto-indexing in autonomous/agent mode - issue 23.
- Added AzureOpenAI provider for Llama.
- Added configuration option for the model in query file/web options in plugins.
- Added removal of temporary index from memory after use.