{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e6b78340-5c44-428c-82db-0aecd6ca7e96",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from mcp101.business_tools import (\n",
    "    calculate_commission,\n",
    "    calculate_total_premium,\n",
    "    filter_policies_by_state,\n",
    "    load_insurance_sales,\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "75352684-14b1-4fba-9751-fa449f7346e9",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Python 3.10.0\n"
     ]
    }
   ],
   "source": [
    "!python --version"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9d3671f-6255-4b7f-a353-4f10c4aa6686",
   "metadata": {},
   "source": [
    "## Step 1: Create the Model Context Protocol Server"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "34a3ed24-8df7-4afd-8bb7-8d1e00fc83ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MCP Server ready! Tools: calculate_profit.\n"
     ]
    }
   ],
   "source": [
    "from fastmcp import FastMCP\n",
    "\n",
    "# Initialize MCP Server\n",
    "mcp = FastMCP(\"BusinessTools\")\n",
    "\n",
    "@mcp.tool()\n",
    "def calculate_profit(revenue: float, expenses: float) -> float:\n",
    "    \"\"\"Calculate profit.\"\"\"\n",
    "    return revenue - expenses\n",
    "\n",
    "print(\"MCP Server ready! Tools: calculate_profit.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c061d2bb-f4fa-4284-9085-5ae2e9ddee55",
   "metadata": {},
   "source": [
    "## Step 2: Test the MCP Tool Locally"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "31892845-8c8d-417a-a441-9893f86183d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Profit: 400\n"
     ]
    }
   ],
   "source": [
    "profit = calculate_profit(1000, 600)\n",
    "print(\"Profit:\", profit)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10f55680-8f86-4b0f-a575-abf2be40aae4",
   "metadata": {},
   "source": [
    "## Step 3: Expand MCP by Adding More Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6bfbbc4d-ecee-4492-a472-b85fca86c44c",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "@mcp.tool()\n",
    "def get_sales_from_csv(file_path: str) -> float:\n",
    "    \"\"\"Read total sales from a CSV file.\"\"\"\n",
    "    df = pd.read_csv(file_path)\n",
    "    return df[\"sales\"].sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1b749f7-f8b3-4ae0-b07f-599040b5ea43",
   "metadata": {},
   "source": [
    "## Step 4: Create the CSV File"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "da41522c-f512-4772-864e-d084010cbbda",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CSV file created successfully!\n"
     ]
    }
   ],
   "source": [
    "csv_data = \"\"\"date,sales\n",
    "2024-02-26,200\n",
    "2024-02-27,450\n",
    "2024-02-28,300\n",
    "\"\"\"\n",
    "\n",
    "# Save to a file in the data folder\n",
    "with open(\"data/sales_data.csv\", \"w\") as file:\n",
    "    file.write(csv_data)\n",
    "\n",
    "print(\"CSV file created successfully!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cff6480d-7f34-42aa-8cbe-b68ba3cedc77",
   "metadata": {},
   "source": [
    "## Step 5: Run get_sales_from_csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ef4b7533-cd81-4d68-9a38-73d778f81c24",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Sales: 950\n"
     ]
    }
   ],
   "source": [
    "sales_total = get_sales_from_csv(\"data/sales_data.csv\")\n",
    "print(\"Total Sales:\", sales_total)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 6: Calculate Commission"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {},
   "execution_count": null,
   "outputs": [],
   "source": [
    "commission = calculate_commission([300, 700, 200], rate=0.1)\n",
    "print('Commission:', commission)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "id": "step-7-load",
   "source": [
    "## Step 7: Load `insurance_sales.csv`"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {},
   "execution_count": null,
   "outputs": [],
   "id": "step-7-code",
   "source": [
    "records = load_insurance_sales('data/insurance_sales.csv')\n",
    "print(f'Total records: {len(records)}')"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {},
   "execution_count": null,
   "outputs": [],
   "id": "step-8-code",
   "source": [
    "total_premium = calculate_total_premium(records)\n",
    "print('Total Premium:', total_premium)"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {},
   "execution_count": null,
   "outputs": [],
   "id": "step-9-code",
   "source": [
    "ca_policies = filter_policies_by_state(records, 'CA')\n",
    "print('CA Policies:', len(ca_policies))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "id": "step-9-explain",
   "source": [
    "These steps show how quickly you can tally premiums and isolate policies by state\u2014handy for reviewing territory performance and carrier payouts."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3aee3fb6-e2ea-4761-b4e1-41d13ef8c507",
   "metadata": {},
   "source": [
    "# \ud83d\ude80 Super Simple Summary: MCP Project Overview\n",
    "\n",
    "## **What We Did** \ud83d\udee0\ufe0f\n",
    "1. **Set up our Python environment**  \n",
    "   - Installed Python 3.10  \n",
    "   - Created a virtual environment (venv)  \n",
    "   - Installed necessary packages (like pandas)  \n",
    "   - Ensured Jupyter Notebook was using the correct environment  \n",
    "\n",
    "2. **Created a simple MCP (Model Context Protocol) server**  \n",
    "   - Used `FastMCP` to set up a tool that can process functions  \n",
    "   - Created a `calculate_profit` function to subtract expenses from revenue  \n",
    "   - Registered that function with MCP  \n",
    "\n",
    "3. **Expanded the MCP tool with more functionality**  \n",
    "   - Added a new function `get_sales_from_csv` to read a CSV file and sum the sales column  \n",
    "   - Used `pandas` to load and process the CSV data  \n",
    "\n",
    "4. **Tested everything** \u2705  \n",
    "   - Created a CSV file inside the notebook  \n",
    "   - Ran our function to sum the sales  \n",
    "   - Verified that the result was correct (`950` total sales)  \n",
    "\n",
    "---\n",
    "\n",
    "## **Why Is This Important?** \ud83d\udd25\n",
    "- **MCP helps structure tools for AI models** \ud83e\udde0  \n",
    "  \u2192 Instead of just writing random functions, MCP lets us build structured \"tools\" that can be reused.  \n",
    "\n",
    "- **We practiced file handling, data processing, and function registration** \ud83d\udcc2  \n",
    "  \u2192 This is a **real-world skill** that applies to **machine learning, AI automation, and software development**.  \n",
    "\n",
    "---\n",
    "\n",
    "## **Next Steps:**\n",
    "Would you like to:  \n",
    "1\ufe0f\u20e3 **Expand this project** (e.g., more data analysis, visualization)?  \n",
    "2\ufe0f\u20e3 **Start a new MCP-based project**?  \n",
    "3\ufe0f\u20e3 **Dive deeper into MCP internals** (how it works under the hood)?  \n",
    "\n",
    "\ud83d\ude80\ud83d\udd25 Let\u2019s build something amazing!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}