Fix LangChain imports and pin versions

app.py

@@ -3,94 +3,111 @@ import re
 import gradio as gr
 
 from langchain_groq import ChatGroq
+from langchain_core.messages import SystemMessage
+
+# ✅ Correct imports for SQLDatabase + Toolkit (HF-safe)
 from langchain_community.utilities import SQLDatabase
 from langchain_community.agent_toolkits import SQLDatabaseToolkit
-from langchain.agents import create_sql_agent
-from langchain_core.messages import SystemMessage
+
+# ✅ Correct import path for create_sql_agent (NOT langchain.agents)
+from langchain_community.agent_toolkits.sql.base import create_sql_agent
 
 
-# --------------------------
-# CONFIG
-# --------------------------
-DB_PATH = "customer_orders.db"
-MODEL_NAME = "meta-llama/llama-4-scout-17b-16e-instruct"
+# -----------------------------
+# Config
+# -----------------------------
+MODEL_NAME = os.getenv("GROQ_MODEL", "meta-llama/llama-3.1-8b-instant")
+GROQ_API_KEY = os.getenv("GROQ_API_KEY")
 
-groq_api_key = os.getenv("GROQ_API_KEY")
-if not groq_api_key:
-    raise ValueError("GROQ_API_KEY not found. Add it in Hugging Face Space → Settings → Secrets.")
+if not GROQ_API_KEY:
+    raise ValueError(
+        "❌ GROQ_API_KEY not found. Add it in Hugging Face: Settings → Secrets → GROQ_API_KEY"
+    )
 
+DB_PATH = "customer_orders.db"  # must be in same Space repo
+DB_URI = f"sqlite:///{DB_PATH}"
+
+if not os.path.exists(DB_PATH):
+    raise FileNotFoundError(
+        f"❌ Database file '{DB_PATH}' not found. Make sure customer_orders.db is committed to the Space repo."
+    )
+
+# -----------------------------
+# LLM
+# -----------------------------
 llm = ChatGroq(
     model=MODEL_NAME,
-    temperature=0,
+    temperature=0,      # deterministic
     max_tokens=300,
-    groq_api_key=groq_api_key
+    groq_api_key=GROQ_API_KEY
 )
 
-# --------------------------
-# LOAD DB + SQL AGENT
-# --------------------------
-db = SQLDatabase.from_uri(f"sqlite:///{DB_PATH}")
+# -----------------------------
+# Database + SQL Agent
+# -----------------------------
+db = SQLDatabase.from_uri(DB_URI)
 toolkit = SQLDatabaseToolkit(db=db, llm=llm)
 
-system_message = SystemMessage(content="""
-You are FoodHub support assistant with database access.
-
-Rules:
-1) Only answer using information from the database.
-2) Only retrieve a SINGLE order when a valid order_id is provided.
-3) If no order_id is provided, ask politely for the Order ID.
-4) Do not reveal bulk order data or all orders.
-5) Replies must be concise, polite and formal.
-""")
+system_message = SystemMessage(content=
+    "You are FoodHub customer support. "
+    "You can query the database safely using SQL tools. "
+    "Rules: (1) Never return bulk data. (2) Only answer for a single order_id. "
+    "(3) If order_id missing, ask for it. (4) Keep replies short and professional."
+)
 
 sql_agent = create_sql_agent(
     llm=llm,
     toolkit=toolkit,
-    verbose=False,
+    verbose=False,  # set True only if you want logs
     system_message=system_message
 )
 
-VALID_ORDER_PATTERN = r"^O\d{5}$"
-
+ORDER_ID_PATTERN = re.compile(r"\bO\d{5}\b", re.IGNORECASE)
 
 def extract_order_id(text: str):
-    match = re.search(r"\bO\d{5}\b", text.upper())
-    return match.group(0) if match else None
+    m = ORDER_ID_PATTERN.search(text or "")
+    return m.group(0).upper() if m else None
 
 
-def chat_fn(user_message, history):
-    order_id = extract_order_id(user_message)
+# -----------------------------
+# Chatbot logic
+# -----------------------------
+def respond(user_message, history):
+    user_message = (user_message or "").strip()
 
-    # If user didn't provide order_id, ask
+    # 1) If order id not provided, ask for it
+    order_id = extract_order_id(user_message)
     if not order_id:
         return "Please provide your Order ID (example: O12488)."
 
-    # Validate order id format
-    if not re.match(VALID_ORDER_PATTERN, order_id):
-        return "Order ID format seems invalid. Please enter like O12488."
+    # 2) Ask SQL agent for that one order only (safe query)
+    # Keep query explicit to avoid the agent doing broad selects
+    agent_query = (
+        f"Retrieve all columns for order_id '{order_id}' from the orders table. "
+        f"Then summarize the order status and payment status in 2-3 short sentences."
+    )
 
-    # Ask SQL agent for that order only
-    query = f"Retrieve all columns for order_id {order_id}"
     try:
-        result = sql_agent.invoke(query)
-        content = result.get("output", str(result))
-
-        # Convert raw output to user-friendly response (simple formatting)
-        # You can customize this further.
-        return f"Here are the details for Order ID {order_id}:\n\n{content}"
-
+        result = sql_agent.invoke({"input": agent_query})
+        # result is usually dict with "output"
+        output_text = result.get("output", str(result))
+        return output_text
     except Exception as e:
-        return f"Sorry, I couldn't fetch that order right now. Error: {e}"
+        return f"Sorry — I couldn’t fetch that order right now. Please try again. (Error: {type(e).__name__})"
 
 
-# --------------------------
-# GRADIO UI
-# --------------------------
+# -----------------------------
+# Gradio UI
+# -----------------------------
 demo = gr.ChatInterface(
-    fn=chat_fn,
+    fn=respond,
     title="FoodHub – AI Powered Food Delivery Chatbot",
-    description="Ask order queries like: 'Where is my order O12488?' or 'Cancel order O12487'.",
-    theme="soft"
+    description="Ask order-related questions. Include your Order ID like O12488.",
+    examples=[
+        "Where is my order? O12488",
+        "What is the payment status of order O12488?",
+        "I want to cancel my order O12487"
+    ],
 )
 
 if __name__ == "__main__":

requirements.txt

@@ -1,8 +1,6 @@
-gradio
-langchain
-langchain-community
-langchain-core
-langchain-groq
-sqlalchemy
-pandas
-numpy
+gradio==4.44.1
+langchain==0.2.17
+langchain-core==0.2.43
+langchain-community==0.2.19
+langchain-groq==0.1.9
+sqlalchemy==2.0.32git add app.py requirements.txt customer_orders.db