18.langgraph-basics.ipynb

{
 "nbformat": 4,
 "nbformat_minor": 0,
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3"
  },
  "language_info": {
   "name": "python"
  }
 },
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "TT_sLhpoqXaf"
   },
   "outputs": [],
   "source": [
    "!pip install -Uqqq pip --progress-bar off\n",
    "!pip install -qqq langchain-groq==0.1.9 --progress-bar off\n",
    "!pip install -qqq langgraph==0.2.4 --progress-bar off"
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "import os\n",
    "import sqlite3\n",
    "from dataclasses import dataclass\n",
    "from datetime import date\n",
    "from enum import Enum, auto\n",
    "from typing import Annotated, List, Literal, Optional, Set, TypedDict\n",
    "\n",
    "from google.colab import userdata\n",
    "from IPython.display import Image, display\n",
    "from langchain_core.messages import HumanMessage, SystemMessage\n",
    "from langchain_core.prompts import PromptTemplate\n",
    "from langchain_core.pydantic_v1 import BaseModel, Field\n",
    "from langchain_core.tools import tool\n",
    "from langchain_groq import ChatGroq\n",
    "from langgraph.checkpoint.memory import MemorySaver\n",
    "from langgraph.graph import END, StateGraph\n",
    "from langgraph.graph.message import add_messages\n",
    "from langgraph.prebuilt import ToolNode\n",
    "\n",
    "MODEL = \"llama3-groq-70b-8192-tool-use-preview\"\n",
    "\n",
    "llm = ChatGroq(temperature=0, model_name=MODEL, api_key=userdata.get(\"GROQ_API_KEY\"))"
   ],
   "metadata": {
    "id": "Aj8xGG_EwNOk"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Habit Tracker"
   ],
   "metadata": {
    "id": "nqbCIAcypuoE"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "DB_NAME = \"habit_tracker.db\"\n",
    "\n",
    "\n",
    "class DayOfWeek(Enum):\n",
    "    MONDAY = auto()\n",
    "    TUESDAY = auto()\n",
    "    WEDNESDAY = auto()\n",
    "    THURSDAY = auto()\n",
    "    FRIDAY = auto()\n",
    "    SATURDAY = auto()\n",
    "    SUNDAY = auto()\n",
    "\n",
    "\n",
    "@dataclass\n",
    "class Habit:\n",
    "    id: int\n",
    "    name: str\n",
    "    repeat_frequency: Set[DayOfWeek]\n",
    "    tags: List[str]\n",
    "\n",
    "\n",
    "@dataclass\n",
    "class DailyHabitEntry:\n",
    "    id: int\n",
    "    name: str\n",
    "    tags: List[str]\n",
    "    is_completed: bool\n",
    "\n",
    "\n",
    "def get_connection():\n",
    "    return sqlite3.connect(DB_NAME)\n",
    "\n",
    "\n",
    "def create_tables():\n",
    "    with get_connection() as conn:\n",
    "        cursor = conn.cursor()\n",
    "        cursor.execute(\n",
    "            \"\"\"\n",
    "            CREATE TABLE IF NOT EXISTS habits (\n",
    "                id INTEGER PRIMARY KEY AUTOINCREMENT,\n",
    "                name TEXT NOT NULL,\n",
    "                repeat_frequency TEXT NOT NULL,\n",
    "                tags TEXT NOT NULL\n",
    "            )\n",
    "        \"\"\"\n",
    "        )\n",
    "        cursor.execute(\n",
    "            \"\"\"\n",
    "            CREATE TABLE IF NOT EXISTS completions (\n",
    "                habit_id INTEGER,\n",
    "                completion_date TEXT,\n",
    "                PRIMARY KEY (habit_id, completion_date),\n",
    "                FOREIGN KEY (habit_id) REFERENCES habits (id)\n",
    "            )\n",
    "        \"\"\"\n",
    "        )\n",
    "        conn.commit()\n",
    "\n",
    "\n",
    "def list_habits() -> List[Habit]:\n",
    "    with get_connection() as conn:\n",
    "        cursor = conn.cursor()\n",
    "        cursor.execute(\"SELECT * FROM habits\")\n",
    "        return [\n",
    "            Habit(\n",
    "                id,\n",
    "                name,\n",
    "                {DayOfWeek[day] for day in freq.split(\",\")},\n",
    "                tags.split(\",\"),\n",
    "            )\n",
    "            for id, name, freq, tags in cursor.fetchall()\n",
    "        ]\n",
    "\n",
    "\n",
    "def habits_for_date(date: date) -> List[DailyHabitEntry]:\n",
    "    weekday = DayOfWeek(date.weekday() + 1).name\n",
    "    with get_connection() as conn:\n",
    "        cursor = conn.cursor()\n",
    "        cursor.execute(\n",
    "            \"\"\"\n",
    "            SELECT h.id, h.name, h.tags, c.completion_date IS NOT NULL as completed\n",
    "            FROM habits h\n",
    "            LEFT JOIN completions c ON h.id = c.habit_id AND c.completion_date = ?\n",
    "            WHERE instr(h.repeat_frequency, ?) > 0\n",
    "        \"\"\",\n",
    "            (date.isoformat(), weekday),\n",
    "        )\n",
    "        return [\n",
    "            DailyHabitEntry(id, name, tags.split(\",\"), bool(completed))\n",
    "            for id, name, tags, completed in cursor.fetchall()\n",
    "        ]\n",
    "\n",
    "\n",
    "def complete_habit(habit_id: int, completion_date: date):\n",
    "    with get_connection() as conn:\n",
    "        cursor = conn.cursor()\n",
    "        cursor.execute(\n",
    "            \"\"\"\n",
    "            INSERT OR REPLACE INTO completions (habit_id, completion_date)\n",
    "            VALUES (?, ?)\n",
    "        \"\"\",\n",
    "            (habit_id, completion_date.isoformat()),\n",
    "        )\n",
    "        conn.commit()\n",
    "\n",
    "\n",
    "def add_habit(name: str, repeat_frequency: Set[DayOfWeek], tags: List[str] = []) -> int:\n",
    "    with get_connection() as conn:\n",
    "        cursor = conn.cursor()\n",
    "        cursor.execute(\n",
    "            \"\"\"\n",
    "            INSERT INTO habits (name, repeat_frequency, tags)\n",
    "            VALUES (?, ?, ?)\n",
    "        \"\"\",\n",
    "            (name, \",\".join(day.name for day in repeat_frequency), \",\".join(tags)),\n",
    "        )\n",
    "        conn.commit()\n",
    "        return cursor.lastrowid\n",
    "\n",
    "\n",
    "def show_habits_for_date(date: date):\n",
    "    print(f\"Habits for {date}:\")\n",
    "    for entry in habits_for_date(date):\n",
    "        status = \"Completed\" if entry.is_completed else \"Not completed\"\n",
    "        print(f\"- {entry.name} (ID: {entry.id}): {status}\")\n",
    "        print(f\"  Tags: {', '.join(entry.tags)}\")\n",
    "\n",
    "\n",
    "create_tables()"
   ],
   "metadata": {
    "id": "eqC9CyQo_-ls"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "add_habit(\"Feed the llamas\", {DayOfWeek.SATURDAY, DayOfWeek.SUNDAY}, [\"diet\"])\n",
    "add_habit(\n",
    "    \"Hit the gym\",\n",
    "    {DayOfWeek.MONDAY, DayOfWeek.WEDNESDAY, DayOfWeek.FRIDAY},\n",
    "    [\"exercise\", \"fitness\"],\n",
    ")"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Sgvs5qBlPbuX",
    "outputId": "fdbbbaee-def8-4a59-b080-4760eb724cb4"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "metadata": {},
     "execution_count": 4
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "list_habits()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "6hqpho6ua2ae",
    "outputId": "ef92ef67-a95a-4886-b7f8-6912da8b4786"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[Habit(id=1, name='Feed the llamas', repeat_frequency={<DayOfWeek.SATURDAY: 6>, <DayOfWeek.SUNDAY: 7>}, tags=['diet']),\n",
       " Habit(id=2, name='Hit the gym', repeat_frequency={<DayOfWeek.MONDAY: 1>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.WEDNESDAY: 3>}, tags=['exercise', 'fitness'])]"
      ]
     },
     "metadata": {},
     "execution_count": 5
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Tools"
   ],
   "metadata": {
    "id": "HlFTV9iUpsYa"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "class CompleteHabitParams(BaseModel):\n",
    "    habit_id: int = Field(description=\"Id (number) of the habit e.g. 2\")\n",
    "    day: int = Field(description=\"Day of the month (1-31) e.g. 18\")\n",
    "    month: int = Field(description=\"Month of the year (1-12) e.g. 3\")\n",
    "    year: int = Field(description=\"Year for which to complete the habit e.g. 2024\")\n",
    "\n",
    "\n",
    "@tool(args_schema=CompleteHabitParams)\n",
    "def complete_habit_tool(habit_id: int, day: int, month: int, year: int):\n",
    "    \"\"\"Completes a habit for a specific date\"\"\"\n",
    "    completion_date = date(year, month, day)\n",
    "    complete_habit(habit_id, completion_date)\n",
    "\n",
    "\n",
    "class HabitsForDateParams(BaseModel):\n",
    "    day: int = Field(description=\"Day of the month (1-31) e.g. 18\")\n",
    "    month: int = Field(description=\"Month of the year (1-12) e.g. 3\")\n",
    "    year: int = Field(description=\"Year for which to complete the habit e.g. 2024\")\n",
    "\n",
    "\n",
    "@tool(args_schema=HabitsForDateParams)\n",
    "def habits_for_date_tool(day: int, month: int, year: int):\n",
    "    \"\"\"Returns a list of habits for a given date\"\"\"\n",
    "    return habits_for_date(date(year, month, day))\n",
    "\n",
    "\n",
    "class AddHabitParams(BaseModel):\n",
    "    name: str = Field(description=\"Name of the habit\")\n",
    "    repeat_frequency: Set[DayOfWeek] = Field(\n",
    "        description=\"Which days of the week the habit should repeat\"\n",
    "    )\n",
    "    tags: Optional[List[str]] = Field(\n",
    "        description=\"Optional list of tags for this habit\"\n",
    "    )\n",
    "\n",
    "\n",
    "@tool(args_schema=AddHabitParams)\n",
    "def add_habit_tool(\n",
    "    name: str, repeat_frequency: Set[DayOfWeek], tags: List[str] = []\n",
    ") -> int:\n",
    "    \"\"\"Adds a new habit with specific repeat frequency. Returns the ID of the habit.\"\"\"\n",
    "    return add_habit(name, repeat_frequency, tags)"
   ],
   "metadata": {
    "id": "UBPrIIx_dnJ8"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "tools = [complete_habit_tool, habits_for_date_tool, add_habit_tool]\n",
    "llm = llm.bind_tools(tools)"
   ],
   "metadata": {
    "id": "CbWPSzdmLUpG"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Build the Graph"
   ],
   "metadata": {
    "id": "0IM_QrD7rRDM"
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## State"
   ],
   "metadata": {
    "id": "6BBgQ5cArS5i"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "class AppState(TypedDict):\n",
    "    messages: Annotated[list, add_messages]\n",
    "    habits: List[Habit]"
   ],
   "metadata": {
    "id": "jdv1LiyDqFJ-"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Nodes"
   ],
   "metadata": {
    "id": "KBqUf7WrrUXW"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "SYSTEM_TEMPLATE = PromptTemplate.from_template(\n",
    "    \"\"\"Current habits\n",
    "{habits}\n",
    "\"\"\"\n",
    ")\n",
    "\n",
    "\n",
    "def update_habits(state: AppState):\n",
    "    return {\"habits\": list_habits()}\n",
    "\n",
    "\n",
    "def call_model(state: AppState):\n",
    "    messages = state[\"messages\"]\n",
    "    first_message = messages[0]\n",
    "    system_message = SystemMessage(SYSTEM_TEMPLATE.format(habits=state[\"habits\"]))\n",
    "    if first_message.type == \"system\":\n",
    "        messages[0] = system_message\n",
    "    else:\n",
    "        messages.insert(0, system_message)\n",
    "    response = llm.invoke(messages)\n",
    "    return {\"messages\": [response]}\n",
    "\n",
    "\n",
    "tool_node = ToolNode(tools)"
   ],
   "metadata": {
    "id": "gargdh3SdrEf"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Edges"
   ],
   "metadata": {
    "id": "OK5OZ69DJfxz"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "def should_continue(state: AppState) -> Literal[\"tools\", END]:\n",
    "    messages = state[\"messages\"]\n",
    "    last_message = messages[-1]\n",
    "    if last_message.tool_calls:\n",
    "        return \"tools\"\n",
    "    return END"
   ],
   "metadata": {
    "id": "nA2IBAs1Jg2v"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Graph"
   ],
   "metadata": {
    "id": "-eeD2W1DtHE6"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "graph = StateGraph(AppState)\n",
    "\n",
    "# nodes of the graph\n",
    "graph.add_node(\"update_habits\", update_habits)\n",
    "graph.add_node(\"agent\", call_model)\n",
    "graph.add_node(\"tools\", tool_node)\n",
    "\n",
    "# edges between nodes\n",
    "graph.add_edge(\"update_habits\", \"agent\")\n",
    "graph.add_conditional_edges(\"agent\", should_continue)\n",
    "graph.add_edge(\"tools\", \"update_habits\")\n",
    "\n",
    "# start node of the graph\n",
    "graph.set_entry_point(\"update_habits\")"
   ],
   "metadata": {
    "id": "ENYse6ymdueN"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "graph.nodes"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "CfrH4h2m3uJE",
    "outputId": "e514049b-7ef9-4aa3-e6ee-539ee758d74e"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "{'update_habits': StateNodeSpec(runnable=update_habits(recurse=True), metadata=None, input=<class '__main__.AppState'>, retry_policy=None),\n",
       " 'agent': StateNodeSpec(runnable=agent(recurse=True), metadata=None, input=<class '__main__.AppState'>, retry_policy=None),\n",
       " 'tools': StateNodeSpec(runnable=tools(recurse=True, tools_by_name={'complete_habit_tool': StructuredTool(name='complete_habit_tool', description='Completes a habit for a specific date', args_schema=<class '__main__.CompleteHabitParams'>, func=<function complete_habit_tool at 0x7947ce8b24d0>), 'habits_for_date_tool': StructuredTool(name='habits_for_date_tool', description='Returns a list of habits for a given date', args_schema=<class '__main__.HabitsForDateParams'>, func=<function habits_for_date_tool at 0x7947ce8b2a70>), 'add_habit_tool': StructuredTool(name='add_habit_tool', description='Adds a new habit with specific repeat frequency. Returns the ID of the habit.', args_schema=<class '__main__.AddHabitParams'>, func=<function add_habit_tool at 0x7947ce8b2cb0>)}, handle_tool_errors=True), metadata=None, input=<class '__main__.AppState'>, retry_policy=None)}"
      ]
     },
     "metadata": {},
     "execution_count": 11
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "graph.edges"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "iQGQMuwO4hU1",
    "outputId": "d809fd1f-4e97-4607-8962-fd71f801ae02"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "{('__start__', 'update_habits'),\n",
       " ('tools', 'update_habits'),\n",
       " ('update_habits', 'agent')}"
      ]
     },
     "metadata": {},
     "execution_count": 12
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "graph.channels"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "-xgnCykJ40J5",
    "outputId": "38ca2ebf-9621-49d4-a935-55c06d3a8af9"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "{'messages': <langgraph.channels.binop.BinaryOperatorAggregate at 0x7947ce89fa60>,\n",
       " 'habits': <langgraph.channels.last_value.LastValue at 0x7947ce89f8b0>}"
      ]
     },
     "metadata": {},
     "execution_count": 13
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Compilation"
   ],
   "metadata": {
    "id": "TodYR7hdtNqd"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "checkpointer = MemorySaver()\n",
    "app = graph.compile(checkpointer=checkpointer)"
   ],
   "metadata": {
    "id": "OAatzQOjd8hY"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "display(Image(app.get_graph().draw_mermaid_png()))"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 320
    },
    "id": "UWOeqN7kPlgB",
    "outputId": "cda703aa-9902-4652-93df-d1862f9fd78e"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/jpeg": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {}
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Chat with the Graph"
   ],
   "metadata": {
    "id": "3occi3pstU0r"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "config = {\"configurable\": {\"thread_id\": 42}}\n",
    "\n",
    "state = app.invoke(\n",
    "    {\"messages\": [HumanMessage(content=\"Complete the workout for August 14th 2024\")]},\n",
    "    config=config,\n",
    ")"
   ],
   "metadata": {
    "id": "WW9N4QBcIOGB"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "state[\"habits\"]"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "kBKSyHSQwVnW",
    "outputId": "82e90362-f1fc-4c02-f370-1c397a176775"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[Habit(id=1, name='Feed the llamas', repeat_frequency={<DayOfWeek.SATURDAY: 6>, <DayOfWeek.SUNDAY: 7>}, tags=['diet']),\n",
       " Habit(id=2, name='Hit the gym', repeat_frequency={<DayOfWeek.MONDAY: 1>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.WEDNESDAY: 3>}, tags=['exercise', 'fitness'])]"
      ]
     },
     "metadata": {},
     "execution_count": 17
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "for m in state[\"messages\"]:\n",
    "    m.pretty_print()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "RPcDpJmvwApE",
    "outputId": "bb9af1c7-30d4-4bb8-c3d5-db7daff1cc62"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "================================\u001b[1m System Message \u001b[0m================================\n",
      "\n",
      "Current habits\n",
      "[Habit(id=1, name='Feed the llamas', repeat_frequency={<DayOfWeek.SATURDAY: 6>, <DayOfWeek.SUNDAY: 7>}, tags=['diet']), Habit(id=2, name='Hit the gym', repeat_frequency={<DayOfWeek.MONDAY: 1>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.WEDNESDAY: 3>}, tags=['exercise', 'fitness'])]\n",
      "\n",
      "================================\u001b[1m Human Message \u001b[0m=================================\n",
      "\n",
      "Complete the workout for August 14th 2024\n",
      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
      "Tool Calls:\n",
      "  complete_habit_tool (call_kkbq)\n",
      " Call ID: call_kkbq\n",
      "  Args:\n",
      "    habit_id: 2\n",
      "    day: 14\n",
      "    month: 8\n",
      "    year: 2024\n",
      "=================================\u001b[1m Tool Message \u001b[0m=================================\n",
      "Name: complete_habit_tool\n",
      "\n",
      "null\n",
      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
      "\n",
      "The workout for August 14th, 2024 has been completed.\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "show_habits_for_date(date(2024, 8, 14))"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ic8GVcYXwh0z",
    "outputId": "0838c455-1233-4cdc-d16a-578dec21828f"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Habits for 2024-08-14:\n",
      "- Hit the gym (ID: 2): Completed\n",
      "  Tags: exercise, fitness\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "print(state[\"messages\"][-1].content)"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "M3QR9RyLeBKt",
    "outputId": "29c486f1-93a9-4d1d-c8c9-b6c76858b1ea"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "The workout for August 14th, 2024 has been completed.\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "show_habits_for_date(date(2024, 8, 14))"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Pc0Mff0Op2xa",
    "outputId": "a3ad1820-7234-471e-c06c-ab74322dca6c"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Habits for 2024-08-14:\n",
      "- Hit the gym (ID: 2): Completed\n",
      "  Tags: exercise, fitness\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "state = app.invoke(\n",
    "    {\n",
    "        \"messages\": [\n",
    "            HumanMessage(\n",
    "                content=\"Add a new habit for Reading a book every weekday #learning\"\n",
    "            )\n",
    "        ]\n",
    "    },\n",
    "    config=config,\n",
    ")"
   ],
   "metadata": {
    "id": "tlVeXWrgIbhv"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "for m in state[\"messages\"]:\n",
    "    m.pretty_print()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "zZyMkh4wwyg8",
    "outputId": "34dead37-4727-4622-f657-a4c3ae0c5405"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "================================\u001b[1m System Message \u001b[0m================================\n",
      "\n",
      "Current habits\n",
      "[Habit(id=1, name='Feed the llamas', repeat_frequency={<DayOfWeek.SATURDAY: 6>, <DayOfWeek.SUNDAY: 7>}, tags=['diet']), Habit(id=2, name='Hit the gym', repeat_frequency={<DayOfWeek.MONDAY: 1>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.WEDNESDAY: 3>}, tags=['exercise', 'fitness']), Habit(id=3, name='Reading a book', repeat_frequency={<DayOfWeek.WEDNESDAY: 3>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.THURSDAY: 4>, <DayOfWeek.MONDAY: 1>, <DayOfWeek.TUESDAY: 2>}, tags=['learning'])]\n",
      "\n",
      "================================\u001b[1m Human Message \u001b[0m=================================\n",
      "\n",
      "Complete the workout for August 14th 2024\n",
      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
      "Tool Calls:\n",
      "  complete_habit_tool (call_kkbq)\n",
      " Call ID: call_kkbq\n",
      "  Args:\n",
      "    habit_id: 2\n",
      "    day: 14\n",
      "    month: 8\n",
      "    year: 2024\n",
      "=================================\u001b[1m Tool Message \u001b[0m=================================\n",
      "Name: complete_habit_tool\n",
      "\n",
      "null\n",
      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
      "\n",
      "The workout for August 14th, 2024 has been completed.\n",
      "================================\u001b[1m Human Message \u001b[0m=================================\n",
      "\n",
      "Add a new habit for Reading a book every weekday #learning\n",
      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
      "Tool Calls:\n",
      "  add_habit_tool (call_q87j)\n",
      " Call ID: call_q87j\n",
      "  Args:\n",
      "    name: Reading a book\n",
      "    repeat_frequency: [1, 2, 3, 4, 5]\n",
      "    tags: ['learning']\n",
      "=================================\u001b[1m Tool Message \u001b[0m=================================\n",
      "Name: add_habit_tool\n",
      "\n",
      "3\n",
      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
      "\n",
      "The new habit for Reading a book every weekday has been added successfully.\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "list_habits()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Kfh7J2oXJ6KZ",
    "outputId": "eed7e3aa-01de-463c-edfb-1c1eb2a690f2"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[Habit(id=1, name='Feed the llamas', repeat_frequency={<DayOfWeek.SATURDAY: 6>, <DayOfWeek.SUNDAY: 7>}, tags=['diet']),\n",
       " Habit(id=2, name='Hit the gym', repeat_frequency={<DayOfWeek.MONDAY: 1>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.WEDNESDAY: 3>}, tags=['exercise', 'fitness']),\n",
       " Habit(id=3, name='Reading a book', repeat_frequency={<DayOfWeek.WEDNESDAY: 3>, <DayOfWeek.FRIDAY: 5>, <DayOfWeek.THURSDAY: 4>, <DayOfWeek.MONDAY: 1>, <DayOfWeek.TUESDAY: 2>}, tags=['learning'])]"
      ]
     },
     "metadata": {},
     "execution_count": 24
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "state = app.invoke(\n",
    "    {\n",
    "        \"messages\": [\n",
    "            HumanMessage(content=\"Show the scheduled habits for 14 August 2024\")\n",
    "        ]\n",
    "    },\n",
    "    config=config,\n",
    ")"
   ],
   "metadata": {
    "id": "sA4sBC6teg7-"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "print(state[\"messages\"][-1].content)"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "K0gMEMaxetF0",
    "outputId": "ac96d1d4-72e6-447a-ecfb-548b96c91ef5"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "The scheduled habits for August 14th, 2024 are:\n",
      "1. Hit the gym (Completed)\n",
      "2. Reading a book (Not Completed)\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "show_habits_for_date(date(2024, 8, 14))"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "qCvDUzY4tF2-",
    "outputId": "c23ca252-79bf-4765-d87a-8c5e0eb13dd6"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Habits for 2024-08-14:\n",
      "- Hit the gym (ID: 2): Completed\n",
      "  Tags: exercise, fitness\n",
      "- Reading a book (ID: 3): Not completed\n",
      "  Tags: learning\n"
     ]
    }
   ]
  }
 ]
}