Big fix

app.py

@@ -1,8 +1,7 @@
 # mcp_groq_gradio.py
-import asyncio
 import gradio as gr
-import os
 import json
+import traceback
 import httpx
 from contextlib import AsyncExitStack
 from typing import Optional
@@ -21,10 +20,9 @@ class MCPGroqClient:
     
     def __init__(self):
         self.session: Optional[ClientSession] = None
-        self.current_model = ""
         self.exit_stack = AsyncExitStack()
+        self.current_model = ""
         self.groq = None
-        self.available_tools = None
 
     async def connect(self):
         """Establish MCP STDIO connection"""
@@ -42,8 +40,6 @@ class MCPGroqClient:
         )
         await self.session.initialize()
 
-        self.available_tools = await self.session.list_tools()
-
     async def stream_response(self, query: str):
         """Handle streaming with failed_generation debugging"""
         messages = [{"role": "user", "content": query}]
@@ -79,8 +75,9 @@ class MCPGroqClient:
 
         except Exception as e:
             # Handle Groq-specific errors
+            traceback.print_exception(e)
             if hasattr(e, "body") and "failed_generation" in e.body:
-                failed_generation = e.response_body["failed_generation"]
+                failed_generation = e.body["failed_generation"]
                 yield f"\n⚠️ Error: Failed to call a function. Invalid generation:\n{failed_generation}"
             else:
                 yield f"\n⚠️ Critical Error: {str(e)}"
@@ -98,18 +95,6 @@ class MCPGroqClient:
                 "parameters": tool.inputSchema
             }
         } for tool in response.tools]
-    
-    async def _get_available_models(self):
-        try:
-            models = self.groq.models.list()
-            return sorted([model.id for model in models.data if model.active])
-        except Exception as e:
-            print(e)
-            return sorted([
-                "llama-3.3-70b-versatile",
-                "llama-3.1-8b-instant",
-                "gemma2-9b-it"
-            ])
 
     async def _process_tool_calls(self, tool_calls, messages):
         for tool in tool_calls:
@@ -169,7 +154,7 @@ def create_interface():
         with gr.Row(visible=False) as chat_row:
             with gr.Column(scale=0.6):
                 chatbot = gr.Chatbot(height=600)
-                input_box = gr.Textbox(placeholder="Type message...", lines=1)
+                input_box = gr.Textbox(placeholder="Type message...")
                 submit_btn = gr.Button("Send", variant="primary")
             
             with gr.Column(scale=0.4):
@@ -178,25 +163,27 @@ def create_interface():
                     interactive=True,
                     visible=False
                 )
-                available_tools = gr.Textbox(
-                    label="Tools Available",
-                    interactive=False,
-                    visible=False
-                )
 
         # Connect Button Logic
         def connect_client(api_key):
             try:
                 # Initialize Groq client with provided API key
                 client.groq = Groq(api_key=api_key, http_client=httpx.Client(verify=False))
-
-                loop = asyncio.new_event_loop()
-                asyncio.set_event_loop(loop)
-                loop.run_until_complete(client.connect())
-
                 models = client.groq.models.list().data
-                active_models = sorted([m.id for m in models if m.active])
-                tools_list = "\n".join([f"• {t.name}: {t.description}" for t in client.available_tools.tools])
+                available_models = sorted([m.id for m in models if m.active])
+                compatible_models = [
+                    "qwen-qwq-32b",
+                    "qwen-2.5-coder-32b",
+                    "qwen-2.5-32b",
+                    "deepseek-r1-distill-qwen-32b",
+                    "deepseek-r1-distill-llama-70b",
+                    "llama-3.3-70b-versatile",
+                    "llama-3.1-8b-instant",
+                    "mixtral-8x7b-32768",
+                    "gemma2-9b-it"
+                ]
+                active_models = sorted([model for model in available_models if model in compatible_models])
+
                 
                 return {
                     connection_status: "Connected ✅",
@@ -206,10 +193,6 @@ def create_interface():
                         value=active_models[0] if active_models else "",
                         visible=True
                     ),
-                    available_tools: gr.update(
-                        value=tools_list,
-                        visible=True
-                    ),
                     connect_btn: gr.update(visible=False),
                     api_key_input: gr.update(interactive=False)
                 }
@@ -222,15 +205,14 @@ def create_interface():
         connect_btn.click(
             connect_client,
             inputs=api_key_input,
-            outputs=[connection_status, chat_row, model_selector, available_tools, connect_btn, api_key_input]
+            outputs=[connection_status, chat_row, model_selector, connect_btn, api_key_input]
         )
 
         # Chat Handling
         async def chat_stream(query, history, selected_model):
-            print(f"Received query: {query}")  # Debugging log
             client.current_model = selected_model
             
-            # Initialize fresh client session if not connected
+            # Initialize fresh client session
             if not client.session:
                 await client.connect()