agent.py

import os

from langchain_openai import ChatOpenAI
from langchain_core.messages import AnyMessage, SystemMessage, HumanMessage, ToolMessage
from langgraph.graph.message import add_messages
from langgraph.graph import MessagesState
from langgraph.graph import StateGraph, START, END
from typing import TypedDict, Annotated, Literal

from langchain_community.tools import BraveSearch  # web search
from langchain_experimental.tools.python.tool import PythonAstREPLTool  # for logic/math problems

from tools import (calculator_basic, datetime_tools, transcribe_audio, transcribe_youtube, query_image, webpage_content, read_excel)
from prompt import system_prompt

from langchain_core.runnables import RunnableConfig  # for LangSmith tracking

# LangSmith to observe the agent
langsmith_api_key = os.getenv("LANGSMITH_API_KEY")
langsmith_tracing = os.getenv("LANGSMITH_TRACING")

def get_llm():
    api_keys = [os.getenv("OPENROUTER_API_KEY"), os.getenv("OPENROUTER_API_KEY_1")]
    last_exception = None
    for key in api_keys:
        if not key:
            continue
        try:
            llm = ChatOpenAI(
                base_url="https://openrouter.ai/api/v1",
                api_key=key,
                model="qwen/qwen3-coder:free",
                temperature=1
            )
            # Optionally, test the key with a trivial call to ensure it's valid
            # llm.invoke([SystemMessage(content="ping")])
            return llm
        except Exception as e:
            last_exception = e
            continue
    raise RuntimeError(f"All OpenRouter API keys failed: {last_exception}")

llm = get_llm()
python_tool = PythonAstREPLTool()
search_tool = BraveSearch.from_api_key(
                    api_key=os.getenv("BRAVE_SEARCH_API"),
                    search_kwargs={"count": 4},  # returns the 4 best results and their URL
                    description="Web search using Brave"
)

community_tools = [search_tool, python_tool]
custom_tools = calculator_basic + datetime_tools + [transcribe_audio, transcribe_youtube, query_image, webpage_content, read_excel]

tools = community_tools + custom_tools
llm_with_tools = llm.bind_tools(tools)

# Prepare tools by name
tools_by_name = {tool.name: tool for tool in tools}

class MessagesState(TypedDict):  # creates the state (is like the agent's memory at any moment)
    messages: Annotated[list[AnyMessage], add_messages]

# LLM node
def llm_call(state: MessagesState):
    return {
        "messages": [
            llm_with_tools.invoke(
                [SystemMessage(content=system_prompt)] + state["messages"]
            )
        ]
    }

# Tool node
def tool_node(state: MessagesState):
    """Executes the tools"""
    
    result = []
    for tool_call in state["messages"][-1].tool_calls:   # gives a list of the tools the LLM decided to call
        tool = tools_by_name[tool_call["name"]]   # look up the actual tool function using a dictionary
        observation = tool.invoke(tool_call["args"])   # executes the tool
        result.append(ToolMessage(content=observation, tool_call_id=tool_call["id"]))   # the result from the tool is added to the memory
    return {"messages": result}   # thanks to add_messages, LangGraph will automatically append the result to the agent's message history

# Conditional edge function to route to the tool node or end based upon whether the LLM made a tool call
def should_continue(state: MessagesState) -> Literal["Action", END]:
    """Decide if we should continue the loop or stop based upon whether the LLM made a tool call"""

    last_message = state["messages"][-1]  # looks at the last message (usually from the LLM)
    
    # If the LLM makes a tool call, then perform an action
    if last_message.tool_calls:
        return "Action"
    # Otherwise, we stop (reply to the user)
    return END

# Build workflow
builder = StateGraph(MessagesState)

# Add nodes
builder.add_node("llm_call", llm_call)
builder.add_node("environment", tool_node)

# Add edges to connect nodes
builder.add_edge(START, "llm_call")
builder.add_conditional_edges(
    "llm_call",
    should_continue,
    {"Action": "environment",  # name returned by should_continue : Name of the next node
     END: END}
)
    # If tool calls -> "Action" -> environment (executes the tool)
    # If no tool calls -> END

builder.add_edge("environment", "llm_call")  # after running the tools go back to the LLM for another round of reasoning

gaia_agent = builder.compile()  # converts my builder into a runnable agent by using gaia_agent.invoke()

# Wrapper class to initialize and call the LangGraph agent with a user question
class LangGraphAgent:
    def __init__(self):
        print("LangGraphAgent initialized.")
        self.question_count = 0  # Track the number of questions processed

    def __call__(self, question: str) -> str:
        # Determine which API key to use based on question count
        # First 50% of questions use OPENROUTER_API_KEY, rest use OPENROUTER_API_KEY_1
        api_key = os.getenv("OPENROUTER_API_KEY") if self.question_count % 2 == 0 else os.getenv("OPENROUTER_API_KEY_1")
        
        # Create a new LLM instance with the selected API key
        current_llm = ChatOpenAI(
            base_url="https://openrouter.ai/api/v1",
            api_key=api_key,
            model="qwen/qwen3-coder:free",
            temperature=1
        )
        
        # Bind tools to the current LLM
        current_llm_with_tools = current_llm.bind_tools(tools)
        
        # Increment question counter for next call
        self.question_count += 1
        
        print(f"Running LangGraphAgent with input: {question[:150]}... (Using API key {self.question_count % 2 + 1})")
        
        # Create a custom LLM node for this specific question
        def custom_llm_call(state: MessagesState):
            return {
                "messages": [
                    current_llm_with_tools.invoke(
                        [SystemMessage(content=system_prompt)] + state["messages"]
                    )
                ]
            }
        
        # Build a new workflow with the custom LLM
        custom_builder = StateGraph(MessagesState)
        custom_builder.add_node("llm_call", custom_llm_call)
        custom_builder.add_node("environment", tool_node)
        custom_builder.add_edge(START, "llm_call")
        custom_builder.add_conditional_edges(
            "llm_call",
            should_continue,
            {"Action": "environment", END: END}
        )
        custom_builder.add_edge("environment", "llm_call")
        custom_agent = custom_builder.compile()
        
        # Prepare the initial state and config
        input_state = {"messages": [HumanMessage(content=question)]}
        config = RunnableConfig(
            config={
                "run_name": "GAIA Agent",
                "tags": ["gaia", "langgraph", "agent"],
                "metadata": {"user_input": question},
                "recursion_limit": 30,
                "tracing": True
            }
        )
        
        # Run the agent
        result = custom_agent.invoke(input_state, config)
        final_response = result["messages"][-1].content
        
        try:
            return final_response.split("FINAL ANSWER:")[-1].strip()
        except Exception:
            print("Could not split on 'FINAL ANSWER:'")
            return final_response