feat: enhance token usage tracking and history management

- Updated HTML assets for improved loading.
- Integrated token usage tracking in chat processing, appending usage details to transcripts.
- Enhanced OpenAIProvider to include usage data in chat completion responses.
- Implemented asynchronous retrieval of recent token usage history.
- Added utility functions for managing transcript files and parsing usage data.
- Updated UI components to reflect changes in usage status handling.
- Ensured consistent usage status definitions across the application.

electron/providers/OpenAIProvider.ts

@@ -5,9 +5,11 @@ import logManager from "@electron/service/logger"
 
 function _transformChunk(chunk: OpenAI.Chat.Completions.ChatCompletionChunk): UniversalChunk {
   const choice = chunk.choices[0];
+  const usage = (chunk as any).usage;
   return {
-    isEnd: choice?.finish_reason != null,
+    isEnd: choice?.finish_reason != null || (chunk.choices.length === 0 && usage != null),
     result: choice?.delta?.content ?? '',
+    usage: usage ?? undefined,
   }
 }
 
@@ -35,6 +37,7 @@ export class OpenAIProvider extends BaseProvider {
         model,
         messages: messages as any,
         stream: true,
+        stream_options: { include_usage: true },
       }, {
         signal: options?.signal,
       });