Update app.py

app.py

@@ -143,23 +143,28 @@ def reasoning_node(state: AgentState) -> AgentState:
     from langchain.schema import HumanMessage, AIMessage
     from langchain.prompts import ChatPromptTemplate
 
-    # Explicitly load the Google API key
+    # Load and verify the Google API key
     GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
     if not GOOGLE_API_KEY:
         raise ValueError("GOOGLE_API_KEY not set in environment variables.")
 
-    # Defensive: Ensure valid message history
-    if not state["history"] or not isinstance(state["history"][-1], HumanMessage):
+    # Ensure history ends with a HumanMessage
+    if not state.get("history") or not isinstance(state["history"][-1], HumanMessage):
+        state["history"] = state.get("history", [])
         state["history"].append(HumanMessage(content="Continue."))
 
-    # Initialize Gemini model
+    # Ensure context is a dictionary
+    if not isinstance(state.get("context"), dict):
+        state["context"] = {}
+
+    # Initialize the Gemini model (via LangChain wrapper)
     llm = ChatGoogleGenerativeAI(
         model="gemini-1.5-flash",
         temperature=0.1,
         google_api_key=GOOGLE_API_KEY
     )
 
-    # Build prompt
+    # Create prompt from messages
     prompt = ChatPromptTemplate.from_messages([
         ("system", (
             "You're an expert problem solver. Analyze the question, select the best tool, "
@@ -170,36 +175,37 @@ def reasoning_node(state: AgentState) -> AgentState:
             "Response Format:\nReasoning: [Your analysis]\nAction: [Tool name OR 'Final Answer']\n"
             "Action Input: [Input for tool OR final response]"
         )),
-        *state['history']
+        *state["history"]
     ])
 
+    # Build and invoke the chain
     chain = prompt | llm
-
-    # Invoke chain with inputs
     response = chain.invoke({
         "context": state["context"],
-        "reasoning": state["reasoning"],
+        "reasoning": state.get("reasoning", ""),
         "question": state["question"]
     })
 
     content = response.content
     reasoning, action, action_input = parse_agent_response(content)
 
-    # Update agent state
-    state['history'].append(AIMessage(content=content))
-    state['reasoning'] += f"\nStep {state['iterations']+1}: {reasoning}"
-    state['iterations'] += 1
+    # Update state
+    state["history"].append(AIMessage(content=content))
+    state["reasoning"] += f"\nStep {state['iterations']+1}: {reasoning}"
+    state["iterations"] += 1
 
+    # Decide next step based on action
     if "final answer" in action.lower():
-        state['history'].append(AIMessage(content=f"FINAL ANSWER: {action_input}"))
+        state["history"].append(AIMessage(content=f"FINAL ANSWER: {action_input}"))
     else:
-        state['context']['current_tool'] = {
+        state["context"]["current_tool"] = {
             "tool": action,
             "input": action_input
         }
 
     return state
 
+
     
 
 def tool_node(state: AgentState) -> AgentState: