app.py

import os
import gradio as gr
import pandas as pd
import traceback
from graph_builder import build_graph
from api_client import fetch_questions, submit_answers

class GaiaAgent:
    def __init__(self):
        self.graph = build_graph()
        self.executor = self.graph.compile()

    def __call__(self, question):
        # Explicitly pass the state dictionary
        result_state = self.executor.invoke({"question": question})
        return result_state.get("final_answer", "No answer generated.")

def run_and_submit_all(profile):
    try:
        space_id = os.getenv("SPACE_ID")
        username = profile if profile else None
        if not username:
            return "❌ Please log in to Hugging Face.", None

        agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
        agent = GaiaAgent()

        questions_data = fetch_questions()
        if not questions_data:
            return "❌ No questions retrieved from GAIA API.", None

        answers_payload = []
        results_log = []

        for item in questions_data:
            task_id = item["task_id"]
            question = item["question"]
            answer = agent(question)
            answers_payload.append({"task_id": task_id, "submitted_answer": answer})
            results_log.append({
                "Task ID": task_id,
                "Question": question,
                "Submitted Answer": answer
            })

        result = submit_answers(username, agent_code, answers_payload)
        final_status = (
            f"✅ Submission Successful!\n"
            f"User: {result.get('username')}\n"
            f"Score: {result.get('score')}%\n"
            f"Correct: {result.get('correct_count')}/{result.get('total_attempted')}\n"
            f"Message: {result.get('message', '')}"
        )
        return final_status, pd.DataFrame(results_log)

    except Exception as e:
        error_message = (
            f"❌ An error occurred:\n{str(e)}\n\n"
            f"Traceback:\n{traceback.format_exc()}"
        )
        return error_message, None

with gr.Blocks() as demo:
    gr.Markdown("# 🤖 GAIA LangGraph Agent")
    gr.Markdown("Log in and run your agent to evaluate on the GAIA benchmark.")

    login_button = gr.LoginButton()
    run_button = gr.Button("🚀 Run Evaluation & Submit All Answers")
    status_output = gr.Textbox(label="Run Status", lines=10)
    results_table = gr.DataFrame(label="Results")

    run_button.click(
        fn=run_and_submit_all,
        inputs=[login_button],
        outputs=[status_output, results_table]
    )

if __name__ == "__main__":
    demo.launch()