updated the context calculation

app/api/endpoints/gemini.py

@@ -54,11 +54,15 @@ async def gemini_chat(
                 "response": response_text
             }
             
+        prompt_content = chat_data.prompt
+        if chat_data.context:
+            prompt_content = f"Context: {chat_data.context}\n\nPrompt: {chat_data.prompt}"
+
         # Using the async generation method provided by the new google-genai library
         # We use await to ensure we don't block the event loop
         response = await client.aio.models.generate_content(
             model="gemini-2.0-flash", 
-            contents=chat_data.prompt
+            contents=prompt_content
         )
         
         # Track usage if valid module
@@ -67,7 +71,7 @@ async def gemini_chat(
             # 1 char ~= 0.25 tokens (rough estimate if exact count not returned)
             # Gemini response usually has usage_metadata
             usage = response.usage_metadata
-            prompt_tokens = usage.prompt_token_count if usage else len(chat_data.prompt) // 4
+            prompt_tokens = usage.prompt_token_count if usage else len(prompt_content) // 4
             completion_tokens = usage.candidates_token_count if usage else len(response.text) // 4
             
             module.ingress_tokens += prompt_tokens