app.py

"""Basic Agent Evaluation Runner – GPT-4.1 edition (HF Spaces)"""
import os
import requests
import gradio as gr
import pandas as pd
from langchain_core.messages import HumanMessage
from agent import build_graph

# --- Constants ------------------------------------------------------------- #
DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"

# --- Agent wrapper --------------------------------------------------------- #
class BasicAgent:
    """LangGraph agent ready for evaluation."""
    def __init__(self):
        print("BasicAgent initialized (using GPT-4.1).")
        # provider="openai" di default
        self.graph = build_graph(provider="openai")

    def __call__(self, question: str) -> str:
        print(f"Agent received question (first 50 chars): {question[:50]}...")
        msgs = [HumanMessage(content=question)]
        result = self.graph.invoke({"messages": msgs})
        answer = result["messages"][-1].content
        return answer[14:]  # strip "FINAL ANSWER: "


# --- Main evaluation logic ------------------------------------------------- #
def run_and_submit_all(profile: gr.OAuthProfile | None):
    # Verifica login
    if not profile:
        return "Please Login to Hugging Face with the button.", None
    username = profile.username
    print(f"User logged in: {username}")

    # Crea agent
    try:
        agent = BasicAgent()
    except Exception as e:
        return f"Error initializing agent: {e}", None

    # Determina URL spazio (link al codice)
    space_id = os.getenv("SPACE_ID", "unknown-space")
    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"

    # --- Fetch domande ------------------------------------------------------ #
    try:
        resp = requests.get(f"{DEFAULT_API_URL}/questions", timeout=15)
        resp.raise_for_status()
        questions_data = resp.json()
    except Exception as e:
        return f"Error fetching questions: {e}", None

    # --- Rispondi con l'agente --------------------------------------------- #
    results_log = []
    answers_payload = []
    for item in questions_data:
        task_id = item.get("task_id")
        q_text = item.get("question")
        try:
            ans = agent(q_text)
            answers_payload.append({"task_id": task_id, "submitted_answer": ans})
            results_log.append({"Task ID": task_id, "Question": q_text, "Submitted Answer": ans})
        except Exception as e:
            results_log.append({"Task ID": task_id, "Question": q_text, "Submitted Answer": f"AGENT ERROR: {e}"})

    # --- Submit ------------------------------------------------------------- #
    submission = {"username": username, "agent_code": agent_code, "answers": answers_payload}
    try:
        resp = requests.post(f"{DEFAULT_API_URL}/submit", json=submission, timeout=60)
        resp.raise_for_status()
        data = resp.json()
        status = (
            f"Submission Successful!\nUser: {data.get('username')}\n"
            f"Overall Score: {data.get('score', 'N/A')}% "
            f"({data.get('correct_count', '?')}/{data.get('total_attempted', '?')} correct)\n"
            f"Message: {data.get('message', 'No message received.')}"
        )
    except Exception as e:
        status = f"Submission Failed: {e}"

    return status, pd.DataFrame(results_log)

# --- Gradio UI ------------------------------------------------------------- #
with gr.Blocks() as demo:
    gr.Markdown("# Basic Agent Evaluation Runner (GPT-4.1)")
    gr.LoginButton()
    run_btn = gr.Button("Run Evaluation & Submit All Answers")
    status_box = gr.Textbox(lines=5, label="Run Status / Submission Result")
    results_tbl = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
    run_btn.click(fn=run_and_submit_all, outputs=[status_box, results_tbl])

if __name__ == "__main__":
    demo.launch(debug=True, share=False)