agent.py

import os
from dotenv import load_dotenv
from langgraph.graph import START, StateGraph, MessagesState
from langgraph.prebuilt import tools_condition
from langgraph.prebuilt import ToolNode
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_groq import ChatGroq
from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint, HuggingFaceEmbeddings
from langchain_community.tools.tavily_search import TavilySearchResults
from langchain_community.document_loaders import WikipediaLoader
from langchain_community.document_loaders import ArxivLoader
from langchain_community.vectorstores import SupabaseVectorStore
from langchain_core.messages import SystemMessage, HumanMessage
from langchain_core.tools import tool
from langchain.tools.retriever import create_retriever_tool


load_dotenv()

@tool
def multiply(a: int, b: int) -> int:
    """Multiply two numbers.
    Args:
        a: first int
        b: second int
    """
    return a * b

@tool
def add(a: int, b: int) -> int:
    """Add two numbers.
    
    Args:
        a: first int
        b: second int
    """
    return a + b

@tool
def subtract(a: int, b: int) -> int:
    """Subtract two numbers.
    
    Args:
        a: first int
        b: second int
    """
    return a - b

@tool
def divide(a: int, b: int) -> int:
    """Divide two numbers.
    
    Args:
        a: first int
        b: second int
    """
    if b == 0:
        raise ValueError("Cannot divide by zero.")
    return a / b

@tool
def modulus(a: int, b: int) -> int:
    """Get the modulus of two numbers.
    
    Args:
        a: first int
        b: second int
    """
    return a % b

@tool
def wiki_search(query: str) -> str:
    """Search Wikipedia for a query and return maximum 2 results.
    
    Args:
        query: The search query."""
    search_docs = WikipediaLoader(query=query, load_max_docs=2).load()
    formatted_search_docs = "\n\n---\n\n".join(
        [
            f'<Document source="{doc.metadata["source"]}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content}\n</Document>'
            for doc in search_docs
        ])
    return {"wiki_results": formatted_search_docs}

@tool
def web_search(query: str) -> str:
    """Search Tavily for a query and return maximum 3 results.
    
    Args:
        query: The search query."""
    search_docs = TavilySearchResults(max_results=3).invoke(query=query)
    formatted_search_docs = "\n\n---\n\n".join(
        [
            f'<Document source="{doc.metadata["source"]}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content}\n</Document>'
            for doc in search_docs
        ])
    return {"web_results": formatted_search_docs}

@tool
def arvix_search(query: str) -> str:
    """Search Arxiv for a query and return maximum 3 result.
    
    Args:
        query: The search query."""
    search_docs = ArxivLoader(query=query, load_max_docs=3).load()
    formatted_search_docs = "\n\n---\n\n".join(
        [
            f'<Document source="{doc.metadata["source"]}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content[:1000]}\n</Document>'
            for doc in search_docs
        ])
    return {"arvix_results": formatted_search_docs}



system_prompt=  '''
You are a helpful assistant with access to tools for answering questions. For each question, think step by step and clearly explain your reasoning. Conclude your response using the following format:

FINAL ANSWER: [YOUR FINAL ANSWER]

Formatting rules for YOUR FINAL ANSWER:

If the answer is a number, write it as digits only. Do not include commas or units (e.g. $, %, etc.) unless explicitly required.

If the answer is a string, avoid articles ("a", "an", "the"), do not use abbreviations (e.g., for cities), and spell out all digits (e.g., "two thousand").

If the answer is a comma-separated list, apply the above rules to each item in the list, depending on whether they are numbers or strings.

Your answer should only begin with "FINAL ANSWER: " followed directly by the formatted result. Nothing else should follow.
'''
sys_msg = SystemMessage(content=system_prompt)

tools = [
    multiply,
    add,
    subtract,
    divide,
    modulus,
    wiki_search,
    web_search,
    arvix_search,
]

# Build graph function
def build_graph(provider: str = "huggingface"):
    """Build the graph"""
    # Load environment variables from .env file
    if provider == "google":
        # Google Gemini
        llm = ChatGoogleGenerativeAI(model="gemini-2.0-flash", temperature=0)
    elif provider == "groq":
        # Groq https://console.groq.com/docs/models
        llm = ChatGroq(model="qwen-qwq-32b", temperature=0) # optional : qwen-qwq-32b gemma2-9b-it
    elif provider == "huggingface":
        llm = ChatHuggingFace(
            llm=HuggingFaceEndpoint(
                repo_id="TinyLlama/TinyLlama-1.1B-Chat-v1.0",
                task="text-generation",  # for chat‐style use “text-generation”
                max_new_tokens=1024,
                do_sample=False,
                repetition_penalty=1.03,
                temperature=0,
            ),
            verbose=True,
        )
    else:
        raise ValueError("Invalid provider. Choose 'google', 'groq' or 'huggingface'.")
    # Bind tools to LLM
    llm_with_tools = llm.bind_tools(tools)

    # Node
    def assistant(state: MessagesState):
        """Assistant node"""
        return {"messages": [llm_with_tools.invoke(state["messages"])]}

    def retriever(state: MessagesState):
        """Retriever node"""
       
        return {"messages": [sys_msg] }
    


    builder = StateGraph(MessagesState)
    builder.add_node("retriever", retriever)
    builder.add_node("assistant", assistant)
    builder.add_node("tools", ToolNode(tools))
    builder.add_edge(START, "retriever")
    builder.add_edge("retriever", "assistant")
    builder.add_conditional_edges(
        "assistant",
        tools_condition,
    )
    builder.add_edge("tools", "assistant"))

    # Compile graph
    return builder.compile()