app.py

from flask import Flask, request, jsonify, send_from_directory
from flask_cors import CORS  # Required for your HTML to talk to this API
from brain import MairaBrain
import os

app = Flask(__name__)
CORS(app)  # Unlocks the neural link for your maira2.html

# --- CONFIGURATION ---
# Using the Llama 3.2 1B model (Smart & Fast)
REPO_ID = "bartowski/Llama-3.2-1B-Instruct-GGUF"
FILENAME = "Llama-3.2-1B-Instruct-Q4_K_M.gguf"

# Initialize Maira's Brain
# This will download the model to the container on first run
maira = MairaBrain(REPO_ID, FILENAME)

# --- ROUTES ---

@app.route("/", methods=["GET"])
def home():
    """Serves the front-end website"""
    # This looks for maira2.html in the same folder as app.py
    return send_from_directory('.', 'maira2.html')

@app.route("/chat", methods=["POST"])
def chat():
    """The API endpoint for both the Website and WhatsApp"""
    try:
        data = request.json
        user_message = data.get("message", "")
        user_id = data.get("user_id", "default_user")

        if not user_message:
            return jsonify({"error": "Message is empty"}), 400

        # Ask the AI for a response
        ai_response = maira.get_response(user_id, user_message)
        
        return jsonify({
            "status": "success",
            "response": ai_response
        })
    
    except Exception as e:
        print(f"CRITICAL ERROR: {e}")
        return jsonify({
            "status": "error", 
            "response": "My neural circuits are lagging. Try again? 🧊"
        }), 500

if __name__ == "__main__":
    # Port 7860 is mandatory for Hugging Face Spaces.
    # It also works perfectly for local testing.
    app.run(host="0.0.0.0", port=7860, debug=False)