app.py

import os
import gradio as gr
import requests
import pandas as pd
import asyncio
import json
from huggingface_hub import login
from smolagents import CodeAgent, InferenceClientModel, DuckDuckGoSearchTool

# --- Constants ---
DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
QUESTIONS_URL = f"{DEFAULT_API_URL}/questions"
SUBMIT_URL = f"{DEFAULT_API_URL}/submit"

# --- Hugging Face Login ---
login(token=os.environ["HUGGINGFACEHUB_API_TOKEN"])

# --- Define Tools ---
search_tool = DuckDuckGoSearchTool()

# --- Main Async Function ---
async def run_and_submit_all(profile: gr.OAuthProfile | None):
    # Initialize Agent
    try:
        agent = CodeAgent(
            tools=[search_tool],
            model=InferenceClientModel(model="mistralai/Magistral-Small-2506"),
            max_steps=5,
            verbosity_level=2
        )
    except Exception as e:
        return f"❌ Agent Initialization Error: {e}", None

    space_id = os.getenv("SPACE_ID", "unknown")
    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"

    # Fetch questions
    try:
        response = requests.get(QUESTIONS_URL, timeout=15)
        response.raise_for_status()
        questions = response.json()
        if not questions:
            return "⚠️ No questions received.", None
    except Exception as e:
        return f"❌ Failed to fetch questions: {e}", None

    answers = []
    logs = []

    for item in questions:
        task_id = item.get("task_id")
        question = item.get("question")
        if not task_id or not question:
            continue

        system_prompt = (
            "You are a general AI assistant. I will ask you a question. "
            "Report your thoughts, and finish your answer with the following template: "
            "FINAL ANSWER: [YOUR FINAL ANSWER]. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings.\n\n"
        )
        full_prompt = system_prompt + f"Question: {question.strip()}"

        try:
            loop = asyncio.get_running_loop()
            result = await loop.run_in_executor(None, lambda: agent(full_prompt))

            if isinstance(result, dict) and "final_answer" in result:
                final_answer = str(result["final_answer"]).strip()
            elif isinstance(result, str):
                if "FINAL ANSWER:" in result:
                    final_answer = result.split("FINAL ANSWER:")[-1].strip()
                else:
                    final_answer = result.strip()
            else:
                final_answer = str(result).strip()

        except Exception as e:
            print(f"[ERROR] Task {task_id} failed: {e}")
            final_answer = f"AGENT ERROR: {e}"

        answers.append({"task_id": task_id, "model_answer": final_answer})
        logs.append({"Task ID": task_id, "Question": question, "Submitted Answer": final_answer})

    valid_answers = [a for a in answers if isinstance(a["task_id"], str) and isinstance(a["model_answer"], str)]

    if not valid_answers:
        return "❌ Agent produced no valid answers.", pd.DataFrame(logs)

    submission = {
        "username": profile.username if profile else "unknown",
        "agent_code": agent_code,
        "answers": valid_answers
    }

    print("[DEBUG] Submitting:\n", json.dumps(submission, indent=2))

    try:
        resp = requests.post(SUBMIT_URL, json=submission, timeout=60)
        resp.raise_for_status()
        result_data = resp.json()

        summary = (
            f"✅ Submission Successful\n"
            f"User: {result_data.get('username')}\n"
            f"Score: {result_data.get('score', 'N/A')}% "
            f"({result_data.get('correct_count')}/{result_data.get('total_attempted')})\n"
            f"Message: {result_data.get('message', 'No message.')}"
        )
        return summary, pd.DataFrame(logs)

    except Exception as e:
        return f"❌ Submission failed: {e}", pd.DataFrame(logs)

# --- Gradio UI ---
with gr.Blocks() as demo:
    gr.Markdown("# 🧠 GAIA Agent Evaluation Interface")
    gr.Markdown("""
    - Log in with your Hugging Face account.
    - Click the button below to run the agent and submit the answers.
    - Wait for the final score to appear.
    """)

    gr.LoginButton()
    run_button = gr.Button("🚀 Run Evaluation & Submit")
    status = gr.Textbox(label="Status", lines=6)
    table = gr.DataFrame(label="Answer Log")

    run_button.click(fn=run_and_submit_all, outputs=[status, table])

# --- Launch ---
if __name__ == "__main__":
    print("Launching Agent Space...")
    demo.launch(debug=True)