agent.py

import os
from langchain_groq import ChatGroq
from langchain.prompts import PromptTemplate
from langgraph.graph import START, StateGraph, MessagesState
from langgraph.prebuilt import ToolNode, tools_condition
from langchain_community.tools.tavily_search import TavilySearchResults
from langchain_core.messages import HumanMessage
from langchain.tools import tool
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.runnables import Runnable
from dotenv import load_dotenv

# Load environment variables from .env
load_dotenv()

# Initialize LLM
def initialize_llm():
    llm = ChatGroq(
        temperature=0,
        model_name="qwen-qwq-32b",
        groq_api_key=os.getenv("GROQ_API_KEY")
    )
    return llm

# Initialize Tavily Search Tool
def initialize_search_tool():
    return TavilySearchResults()

# Weather tool
def get_weather(location: str, search_tool: TavilySearchResults = None) -> str:
    if search_tool is None:
        search_tool = initialize_search_tool()
    query = f"current weather in {location}"
    return search_tool.run(query)

# Recommendation chain
def initialize_recommendation_chain(llm: ChatGroq) -> Runnable:
    recommendation_prompt = ChatPromptTemplate.from_template("""
    You are a helpful assistant that gives weather-based advice.
    
    Given the current weather condition: "{weather_condition}", provide:
    1. Clothing or activity recommendations suited for this weather.
    2. At least one health tip to stay safe or comfortable in this condition.

    Be concise and clear.
    """)
    return recommendation_prompt | llm

def get_recommendation(weather_condition: str, recommendation_chain: Runnable = None) -> str:
    if recommendation_chain is None:
        llm = initialize_llm()
        recommendation_chain = initialize_recommendation_chain(llm)
    return recommendation_chain.invoke({"weather_condition": weather_condition})

# Math tools
@tool
def add(x: int, y: int) -> int:
    return x + y

@tool
def subtract(x: int, y: int) -> int:
    return x - y

@tool
def multiply(x: int, y: int) -> int:
    return x * y

@tool
def divide(x: int, y: int) -> float:
    if y == 0:
        raise ValueError("Cannot divide by zero.")
    return x / y

@tool
def square(x: int) -> int:
    return x * x

@tool
def cube(x: int) -> int:
    return x * x * x

@tool
def power(x: int, y: int) -> int:
    return x ** y

@tool
def factorial(n: int) -> int:
    if n < 0:
        raise ValueError("Factorial is not defined for negative numbers.")
    if n == 0 or n == 1:
        return 1
    result = 1
    for i in range(2, n + 1):
        result *= i
    return result

@tool
def mean(numbers: list) -> float:
    if not numbers:
        raise ValueError("The list is empty.")
    return sum(numbers) / len(numbers)

@tool
def standard_deviation(numbers: list) -> float:
    if not numbers:
        raise ValueError("The list is empty.")
    mean_value = mean(numbers)
    variance = sum((x - mean_value) ** 2 for x in numbers) / len(numbers)
    return variance ** 0.5

# Build the LangGraph
def build_graph():
    llm = initialize_llm()
    search_tool = initialize_search_tool()
    recommendation_chain = initialize_recommendation_chain(llm)

    @tool
    def weather_tool(location: str) -> str:
        return get_weather(location, search_tool)

    @tool
    def recommendation_tool(weather_condition: str) -> str:
        return get_recommendation(weather_condition, recommendation_chain)

    tools = [weather_tool, recommendation_tool, add, subtract, multiply, divide, square, cube, power, factorial, mean, standard_deviation]

    llm_with_tools = llm.bind_tools(tools)

    def assistant(state: MessagesState):
        print("Entering assistant node...")
        response = llm_with_tools.invoke(state["messages"])
        print(f"Assistant says: {response.content}")
        return {"messages": [response]}

    builder = StateGraph(MessagesState)
    builder.add_node("assistant", assistant)
    builder.add_node("tools", ToolNode(tools))
    builder.set_entry_point("assistant")
    builder.add_conditional_edges("assistant", tools_condition)
    builder.add_edge("tools", "assistant")
    return builder.compile()

if __name__ == "__main__":
    graph = build_graph()
    question = "What is the factorial of 6 and can you also tell me the weather in Paris?"
    messages = [HumanMessage(content=question)]
    result = graph.invoke({"messages": messages})
    for msg in result["messages"]:
        msg.pretty_print()