app.py

import os
import json
import operator
import streamlit as st
import tempfile
from typing import TypedDict, Annotated, Sequence
from langchain_openai import ChatOpenAI
from langchain_core.tools import tool
from langchain_core.messages import HumanMessage, AIMessage
from langchain_core.utils.function_calling import convert_to_openai_tool
from langgraph.graph import StateGraph, END

# Environment Setup
os.environ['OPENAI_API_KEY'] = os.getenv("OPENAI_API_KEY")

# Model Initialization
model = ChatOpenAI(temperature=0)

# Define the tool
@tool
def multiply(first_number: int, second_number: int):
    """Multiplies two numbers together and returns the result."""
    return first_number * second_number

# Bind tool to model
model_with_tools = model.bind(tools=[convert_to_openai_tool(multiply)])

# State Setup
class AgentState(TypedDict):
    messages: Annotated[Sequence, operator.add]

graph = StateGraph(AgentState)

# Model Invocation
def invoke_model(state):
    question = state['messages'][-1].content
    response = model_with_tools.invoke(question)
    return {"messages": [response]}

graph.add_node("agent", invoke_model)

# Tool Invocation
def invoke_tool(state):
    tool_calls = state['messages'][-1].additional_kwargs.get("tool_calls", [])
    tool_results = []

    for tool_call in tool_calls:
        if tool_call["function"]["name"] == "multiply":
            arguments = json.loads(tool_call["function"]["arguments"])
            result = multiply.invoke(arguments)
            tool_results.append(
                AIMessage(content=f"Tool Result: {result}", additional_kwargs={"tool_call_id": tool_call["id"]})
            )
    return {"messages": tool_results or [AIMessage(content="No valid tool input provided.")]}

graph.add_node("tool", invoke_tool)

# Fixed Router Node
def router(state):
    """
    Route to tool if tool calls exist; otherwise END the workflow.
    """
    tool_calls = state['messages'][-1].additional_kwargs.get("tool_calls", [])
    if tool_calls:
        return {"__state__": "tool"}  # Direct routing to tool
    else:
        return {"__state__": END}  # End workflow if no tool calls

graph.add_node("router", router)

# Add explicit edges
graph.add_edge("agent", "router")
graph.add_conditional_edges(
    "router",
    lambda state: "tool" if state['messages'][-1].additional_kwargs.get("tool_calls") else END,
    {"tool": "tool", END: END}
)
graph.add_edge("tool", END)

# Compile the graph
graph.set_entry_point("agent")
app_graph = graph.compile()

# Save graph visualization with xray for visibility
with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as tmpfile:
    graph_viz = app_graph.get_graph(xray=True)
    tmpfile.write(graph_viz.draw_mermaid_png())
    graph_image_path = tmpfile.name

# Streamlit Interface
st.title("Simple Tool Calling Demo")
st.image(graph_image_path, caption="Workflow Visualization")

tab1, tab2 = st.tabs(["Try Multiplication", "Ask General Queries"])

with tab1:
    st.subheader("Try Multiplication")
    first_number = st.number_input("First Number", value=0, step=1)
    second_number = st.number_input("Second Number", value=0, step=1)

    if st.button("Multiply"):
        question = f"What is {first_number} * {second_number}?"
        try:
            output = app_graph.invoke({"messages": [HumanMessage(content=question)]})
            st.success(output['messages'][-1].content)
        except Exception as e:
            st.error(f"Error: {e}")

with tab2:
    st.subheader("General Query")
    user_input = st.text_input("Enter your question here")

    if st.button("Submit"):
        try:
            result = app_graph.invoke({"messages": [HumanMessage(content=user_input)]})
            st.success(result['messages'][-1].content)
        except Exception as e:
            st.error(f"Error: {e}")

# Sidebar for References
st.sidebar.title("Reference:")
st.sidebar.markdown("1. [LangGraph Tool Calling](https://github.com/aritrasen87/LLM_RAG_Model_Deployment/blob/main/LangGraph_02_ToolCalling.ipynb)")