Restore HuggingFace imports and create hybrid AI brain with HF primary + Groq fallback

agent.py

@@ -24,7 +24,10 @@ from langgraph.graph import StateGraph, END
 from langgraph.graph.message import add_messages
 from typing_extensions import TypedDict
 
-# Groq imports for free AI model
+# HuggingFace imports
+from huggingface_hub import InferenceClient
+
+# Groq imports for fallback
 from groq import Groq
 
 # Utils system imports
@@ -68,33 +71,55 @@ class AgentState(TypedDict):
 # =============================================================================
 
 class LangChainQwen3Brain:
-    """AI Brain using LangChain + Groq free models"""
+    """AI Brain using LangChain + HuggingFace with Groq fallback"""
     
     def __init__(self):
-        # Use Groq instead of HuggingFace
-        self.client = Groq(
+        # Primary: HuggingFace
+        self.hf_client = InferenceClient(
+            provider="auto", 
+            api_key=os.environ.get("HF_TOKEN", "")
+        )
+        self.hf_model = "Qwen/Qwen3-8B"
+        
+        # Fallback: Groq  
+        self.groq_client = Groq(
             api_key=os.environ.get("GROQ_API_KEY", "")
         )
-        self.model_name = "llama3-8b-8192"  # Free Groq model
+        self.groq_model = "llama3-8b-8192"
         
         # Setup parsers
         self.json_parser = JsonOutputParser()
         self.str_parser = StrOutputParser()
         
-        print("🧠 LangChain Groq Brain initialized")
+        print("🧠 LangChain Hybrid Brain initialized (HF + Groq fallback)")
     
     def _invoke_model(self, messages: List[Dict[str, str]]) -> str:
-        """Invoke model with messages"""
+        """Invoke model with messages - try HF first, fallback to Groq"""
+        
+        # Try HuggingFace first
         try:
-            completion = self.client.chat.completions.create(
-                model=self.model_name,
+            completion = self.hf_client.chat.completions.create(
+                model=self.hf_model,
                 messages=messages,
                 max_tokens=2048,
                 temperature=0.7
             )
             return completion.choices[0].message.content
-        except Exception as e:
-            return f"AI Error: {str(e)}"
+        except Exception as hf_error:
+            print(f"⚠️ HuggingFace failed: {str(hf_error)[:100]}...")
+            print("🔄 Falling back to Groq...")
+            
+            # Fallback to Groq
+            try:
+                completion = self.groq_client.chat.completions.create(
+                    model=self.groq_model,
+                    messages=messages,
+                    max_tokens=2048,
+                    temperature=0.7
+                )
+                return completion.choices[0].message.content
+            except Exception as groq_error:
+                return f"AI Error: Both HF ({str(hf_error)[:50]}) and Groq ({str(groq_error)[:50]}) failed"
     
     def analyze_question(self, question: str, task_id: str = "") -> Dict[str, Any]:
         """AI analyzes question and decides approach"""
@@ -386,7 +411,7 @@ class LangGraphUtilsAgent:
         self.ai_brain = ai_brain
         
         print("🤖 LangGraph Utils Agent initialized!")
-        print("🧠 AI Brain: Groq free models with LangChain")
+        print("🧠 AI Brain: LangChain + HuggingFace with Groq fallback")
         print("🔧 Tools: YouTube, Image OCR, Audio Transcript, Wikipedia, File Reader, Text Processor")
         print("⚡ Features: AI-driven routing, Smart tool selection, Multimodal processing")