Update router.py

router.py

@@ -7,14 +7,20 @@ from llm_clients import (
 
 from memory import save_message, load_memory
 from search_tool import search_web
+from rag_engine import rag_response
+
 import requests
 
 
-# Optional: image microservice URL
+# 🔥 Image microservice endpoint
 IMAGE_SPACE_URL = "https://your-image-space.hf.space/generate"
 
+# 🔥 Simple in-memory cache
+response_cache = {}
+
 
 def build_messages(system_prompt, memory, user_prompt):
+
     messages = []
 
     if system_prompt:
@@ -27,6 +33,7 @@ def build_messages(system_prompt, memory, user_prompt):
 
 
 def call_image_microservice(prompt):
+
     try:
         response = requests.post(
             IMAGE_SPACE_URL,
@@ -40,51 +47,73 @@ def call_image_microservice(prompt):
 
 def route_request(prompt, user_id):
 
+    cache_key = f"{user_id}:{prompt}"
+
+    # ==========================
+    # CACHE CHECK
+    # ==========================
+    if cache_key in response_cache:
+        return {"response": response_cache[cache_key]}
+
     # ==========================
-    # Image Command
+    # IMAGE COMMAND
     # ==========================
     if prompt.startswith("/image"):
         clean_prompt = prompt.replace("/image", "").strip()
         return call_image_microservice(clean_prompt)
 
     # ==========================
-    # Load Memory
+    # RAG QUICK RESPONSE
+    # ==========================
+    rag_answer = rag_response(prompt)
+
+    if rag_answer:
+        response_cache[cache_key] = rag_answer
+        return {"response": rag_answer}
+
+    # ==========================
+    # LOAD MEMORY
     # ==========================
     memory = load_memory(user_id)
 
     # ==========================
-    # AI Classification
+    # AI CLASSIFICATION
     # ==========================
     classification = classify_prompt(prompt)
 
     intent = classification.get("intent", "chat")
     needs_search = classification.get("needs_search", False)
 
-    system_prompt = "You are ZXAI, a helpful advanced AI assistant."
+    system_prompt = "You are ZXAI, a powerful advanced AI assistant."
 
     # ==========================
-    # Greeting (fast return)
+    # GREETING FAST PATH
     # ==========================
     if intent == "greeting":
         response = "Hello 👋 I am ZXAI. How can I assist you today?"
+
         save_message(user_id, "user", prompt)
         save_message(user_id, "assistant", response)
+
+        response_cache[cache_key] = response
         return {"response": response}
 
     # ==========================
-    # Reasoning → Gemini
+    # REASONING → GEMINI
     # ==========================
     if intent == "reasoning":
+
         messages = build_messages(system_prompt, memory, prompt)
         response = call_gemini(messages)
 
         save_message(user_id, "user", prompt)
         save_message(user_id, "assistant", response)
 
+        response_cache[cache_key] = response
         return {"response": response}
 
     # ==========================
-    # Live Data
+    # LIVE DATA / SEARCH
     # ==========================
     if intent == "live_data" or needs_search:
 
@@ -97,7 +126,7 @@ User Question:
 Web Data:
 {web_data}
 
-Use the web data if relevant.
+Use web data if helpful.
 """
 
         messages = build_messages(system_prompt, memory, enriched_prompt)
@@ -112,10 +141,11 @@ Use the web data if relevant.
         save_message(user_id, "user", prompt)
         save_message(user_id, "assistant", final_answer)
 
+        response_cache[cache_key] = final_answer
         return {"response": final_answer}
 
     # ==========================
-    # Default Chat → Llama
+    # DEFAULT CHAT → LLAMA
     # ==========================
     messages = build_messages(system_prompt, memory, prompt)
 
@@ -124,5 +154,6 @@ Use the web data if relevant.
     save_message(user_id, "user", prompt)
     save_message(user_id, "assistant", response)
 
+    response_cache[cache_key] = response
+
     return {"response": response}
-