Update app.py

app.py

@@ -24,60 +24,54 @@ except importlib.metadata.PackageNotFoundError:
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# --- Basic Agent Definition ---
-# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
-class AgentState(TypedDict):
-    # Keep track of the full “chat history” so the LLM sees it each time
-    messages: list[str]
-    # These two fields are _optional_—we won’t even use them directly in our code,
-    # but the ReAct agent will populate them when it calls a tool.
-    tool_name: str       # e.g. "ocr_image" or "web_search"
-    tool_input: str      # whatever arguments you passed into the tool
-
-# ─── 2) Instantiate your LLM (ChatOpenAI) ───
-llm = ChatOpenAI(model_name="gpt-4.1-mini", temperature=0.0)
 
-# ─── 3) Wrap your three tools into a single ToolNode ───
-# ToolNode lets LangGraph know “these are the only tool functions the agent may call”
+llm = ChatOpenAI(model_name="gpt-4.1-mini", temperature=0.0)
 tool_node = ToolNode([ocr_image, parse_excel, web_search])
 
-# ─── 4) Use create_react_agent to build a ReAct‐style agent for you ───
-# This single “agent” node will:
-#   • Take the entire AgentState (including messages),
-#   • Look at state["messages"], decide if it needs to call a tool,
-#   • If so, emit {"tool": "<tool_name>", "input": "<tool_input>"},
-#     and then feed results back into the LLM automatically,
-#   • If not, emit a final answer as plain text.
-agent = create_react_agent(llm, tool_node)
-
-# ─── 5) Build a graph with exactly two edges, just like the tutorial ───
-graph = StateGraph(AgentState)
+agent = create_react_agent(
+    model=llm,
+    tools=tool_node
+    # (Use default prompt/state_schema; do NOT pass your own TypedDict.)
+)
+
+# ─── 4) Build a graph that simply wires START → "agent" → END ───
+graph = StateGraph(dict)     # We’ll use plain dicts instead of a custom TypedDict
 graph.add_node("agent", agent)
 
-# 5.a) Whenever user input arrives, send it into the “agent” node:
+# 4.a) Whenever user input arrives, send it straight into the agent
 graph.add_edge(START, "agent")
 
-# 5.b) Once “agent” produces its final text (not a tool call), go to END:
+# 4.b) Once the agent returns a final answer, go to END
 graph.add_edge("agent", END)
 
-# 5.c) Compile so we can call .invoke(...) at runtime
+# 4.c) Compile so we can call `.invoke()` at runtime
 compiled_graph = graph.compile()
 
-# ─── 6) Define a simple function Gradio (and the “submit all” loop) can call ───
+# ─── 5) Define `respond_to_input` to call `compiled_graph.invoke` ───
 def respond_to_input(user_input: str) -> str:
-    initial_state: AgentState = {
-        "messages": [],  # No history on the first turn
-        "tool_name": "",
-        "tool_input": ""
+    """
+    We must supply `{"messages": [ { "role": "user", "content": ... } ]}`
+    because create_react_agent expects `state["messages"]` to be a list of
+    {role,content} dicts, not plain strings.
+    """
+    initial_state: Dict[str, Any] = {
+        "messages": [
+            {"role": "user", "content": user_input}
+        ]
     }
-    # In v0.4.7, use .invoke(...) on the compiled graph
-    final_state = compiled_graph.invoke(initial_state, user_input)
-    # `create_react_agent` always appends its LLM text into state["messages"].
-    # The last entry of “ASSISTANT: <answer>” is the final answer.
-    last_line = final_state["messages"][-1]
-    # If the agent prefixes with “ASSISTANT: ”, strip it off.
-    return last_line.replace("ASSISTANT: ", "", 1)
-
+    # In langgraph 0.4.7, the compiled graph uses .invoke(...)
+    final_state = compiled_graph.invoke(initial_state)
+    # The default agent appends its own assistant messages to state["messages"].
+    # Extract the **last** message where role == "assistant".
+    assistant_messages = [
+        msg["content"]
+        for msg in final_state["messages"]
+        if msg.get("role") == "assistant"
+    ]
+    if not assistant_messages:
+        return "❗️Agent did not return any assistant message."
+    # Return the very last assistant “content”
+    return assistant_messages[-1]
 
 class BasicAgent:
     def __init__(self):