Update app.py

app.py

@@ -1,8 +1,13 @@
+""" Basic Agent Evaluation Runner"""
 import os
+import inspect
 import gradio as gr
 import requests
-import inspect
 import pandas as pd
+from langchain_core.messages import HumanMessage
+from agent import build_graph
+
+
 
 # (Keep Constants as is)
 # --- Constants ---
@@ -10,48 +15,22 @@ DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
 # --- Basic Agent Definition ---
 # ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
-from agent import tools  # import tools from your agent.py
-from langchain.schema import HumanMessage
-from langchain_community.chat_models import ChatOllama
+
 
 class BasicAgent:
+    """A langgraph agent."""
     def __init__(self):
-        print("Custom BasicAgent initialized with tools.")
-        self.tools = {tool.name: tool for tool in tools}
-        self.llm = ChatOllama(model="llama3", temperature=0)  # match your agent.py setup
+        print("BasicAgent initialized.")
+        self.graph = build_graph()
 
     def __call__(self, question: str) -> str:
         print(f"Agent received question (first 50 chars): {question[:50]}...")
+        # Wrap the question in a HumanMessage from langchain_core
+        messages = [HumanMessage(content=question)]
+        messages = self.graph.invoke({"messages": messages})
+        answer = messages['messages'][-1].content
+        return answer[14:]
 
-        # Attempt to route to the best matching tool
-        for tool in self.tools.values():
-            inputs = tool.inputs
-            try:
-                # Try calling the tool with the question as input
-                if list(inputs.keys()) == ["expression"]:
-                    return tool.forward(question)
-                elif list(inputs.keys()) == ["query"]:
-                    return tool.forward(question)
-                elif list(inputs.keys()) == ["topic"]:
-                    return tool.forward(question)
-                elif list(inputs.keys()) == ["code"]:
-                    return tool.forward(question)
-                elif list(inputs.keys()) == ["question"]:
-                    return tool.forward(question)
-                elif set(inputs.keys()) == {"context", "query"}:
-                    return tool.forward(context=question, query=question)
-                elif set(inputs.keys()) == {"story", "question"}:
-                    return tool.forward(story=question, question=question)
-            except Exception as e:
-                print(f"Tool '{tool.name}' failed: {e}")
-                continue
-
-        # Fallback: use LLM directly
-        try:
-            response = self.llm.generate([HumanMessage(content=question)])
-            return response.content.strip()
-        except Exception as e:
-            return f"Fallback LLM error: {e}"
 
 def run_and_submit_all( profile: gr.OAuthProfile | None):
     """