Feat: Implement token tracking, soft delete, and Admin UI improvements

app/api/endpoints/openai.py

@@ -1,5 +1,8 @@
 from fastapi import APIRouter, Depends, Request
-from app.api.deps import get_api_key
+from app.api.deps import get_api_key, get_current_module
+from app.models.module import Module
+from sqlalchemy.orm import Session
+from app.core.database import get_db
 from app.core.limiter import limiter
 from app.core.config import settings
 from pydantic import BaseModel
@@ -22,14 +25,27 @@ if settings.OPENAI_API_KEY and settings.OPENAI_API_KEY != "your-openai-api-key":
 async def openai_chat(
     request: Request,
     chat_data: LLMRequest,
-    api_key: str = Depends(get_api_key)
+    api_key: str = Depends(get_api_key),
+    module: Module = Depends(get_current_module),
+    db: Session = Depends(get_db)
 ):
     try:
         if not client:
+             # Mock response
+             response_text = f"MOCK: OpenAI response to '{chat_data.prompt}'"
+             if module:
+                 # Estimate tokens for mock
+                 prompt_tokens = len(chat_data.prompt) // 4
+                 completion_tokens = len(response_text) // 4
+                 module.ingress_tokens += prompt_tokens
+                 module.egress_tokens += completion_tokens
+                 module.total_tokens += (prompt_tokens + completion_tokens)
+                 db.commit()
+
              return {
                 "status": "mock",
                 "model": "openai",
-                "response": f"MOCK: OpenAI response to '{chat_data.prompt}'"
+                "response": response_text
             }
 
         # Perform Async call to OpenAI
@@ -37,6 +53,16 @@ async def openai_chat(
             model="gpt-3.5-turbo",
             messages=[{"role": "user", "content": chat_data.prompt}]
         )
+        
+        # Track usage
+        if module:
+            usage = response.usage
+            if usage:
+                module.ingress_tokens += usage.prompt_tokens
+                module.egress_tokens += usage.completion_tokens
+                module.total_tokens += usage.total_tokens
+                db.commit()
+
         return {
             "status": "success",
             "model": "openai",