{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "18772359",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ok\n"
     ]
    }
   ],
   "source": [
    "print(\"ok\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "68b7c55e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from dotenv import load_dotenv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a0255cca",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_groq import ChatGroq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a73c057d",
   "metadata": {},
   "outputs": [],
   "source": [
    "llm=ChatGroq(model=\"deepseek-r1-distill-llama-70b\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ec4c38f3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AIMessage(content='<think>\\n\\n</think>\\n\\nHello! How can I assist you today? 😊', additional_kwargs={}, response_metadata={'token_usage': {'completion_tokens': 16, 'prompt_tokens': 4, 'total_tokens': 20, 'completion_time': 0.084725173, 'prompt_time': 6.0159e-05, 'queue_time': 0.053524741, 'total_time': 0.084785332}, 'model_name': 'deepseek-r1-distill-llama-70b', 'system_fingerprint': 'fp_1bbe7845ec', 'finish_reason': 'stop', 'logprobs': None}, id='run--8b091ea3-0a7d-4058-9e8b-2a68094cc078-0', usage_metadata={'input_tokens': 4, 'output_tokens': 16, 'total_tokens': 20})"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "llm.invoke(\"hi\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "0215d359",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.tools import tool"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d81ac345",
   "metadata": {},
   "outputs": [],
   "source": [
    "@tool\n",
    "def multiply(a: int, b: int) -> int:\n",
    "    \"\"\"\n",
    "    Multiply two integers.\n",
    "\n",
    "    Args:\n",
    "        a (int): The first integer.\n",
    "        b (int): The second integer.\n",
    "\n",
    "    Returns:\n",
    "        int: The product of a and b.\n",
    "    \"\"\"\n",
    "    return a * b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1f79ae97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "StructuredTool(name='multiply', description='Multiply two integers.\\n\\nArgs:\\n    a (int): The first integer.\\n    b (int): The second integer.\\n\\nReturns:\\n    int: The product of a and b.', args_schema=<class 'langchain_core.utils.pydantic.multiply'>, func=<function multiply at 0x000002DFC40C3A30>)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multiply"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2927babf",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.tools import StructuredTool"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "03483480",
   "metadata": {},
   "outputs": [],
   "source": [
    "class WeatherInput(BaseModel):\n",
    "    city: str"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "28d0a2ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_weather(city: str) -> str:\n",
    "    \"\"\"\n",
    "    Get the weather for a given city.\n",
    "\n",
    "    Args:\n",
    "        city (str): The name of the city.\n",
    "\n",
    "    Returns:\n",
    "        str: A string describing the weather in the city.\n",
    "    \"\"\"\n",
    "    return f\"The weather in {city} is sunny.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b45850a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "weather_tool = StructuredTool.from_function(\n",
    "    func=get_weather,\n",
    "    name=\"get_weather\",\n",
    "    description=\"Fetches real-time weather data for a city\",\n",
    "    args_schema=WeatherInput,  \n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e36ee4d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "class WeatherInput(BaseModel):\n",
    "    city: str = Field(..., description=\"City name\")\n",
    "    units: str = Field(\"metric\", description=\"metric or imperial\")\n",
    "\n",
    "class GetWeatherTool(StructuredTool):\n",
    "    name: ClassVar[str] = \"get_weather\"           \n",
    "    description: ClassVar[str] = (\n",
    "        \"Fetches weather data for a city\"\n",
    "    )\n",
    "    args_schema: ClassVar[Type[BaseModel]] = WeatherInput\n",
    "\n",
    "    def _run(self, city: str, units: str = \"metric\") -> str:\n",
    "        return get_weather(city, units)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "647b5794",
   "metadata": {},
   "source": [
    "## SmartTrip AI API Key Diagnostics\n",
    "\n",
    "This section checks whether each external service key works without printing secret values. It tests weather, places, Tavily, exchange-rate, the configured LLM, and finally the full LangGraph agent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4a2fe847",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Project root: f:\\hugging_face_host_projects\\smarttrip_ai\n",
      "\n",
      "Environment key presence:\n",
      "MODEL_PROVIDER           SET\n",
      "MODEL_NAME               SET\n",
      "MODEL_BASE_URL           SET\n",
      "OLLAMA_API_KEY           SET\n",
      "OPENWEATHERMAP_API_KEY   SET\n",
      "FOURSQUARE_API_KEY       SET\n",
      "TAVILY_API_KEY           SET\n",
      "EXCHANGE_RATE_API_KEY    SET\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import sys\n",
    "from pathlib import Path\n",
    "\n",
    "PROJECT_ROOT = Path.cwd()\n",
    "if PROJECT_ROOT.name == \"notebook\":\n",
    "    PROJECT_ROOT = PROJECT_ROOT.parent\n",
    "\n",
    "sys.path.insert(0, str(PROJECT_ROOT))\n",
    "os.chdir(PROJECT_ROOT)\n",
    "\n",
    "from dotenv import load_dotenv\n",
    "load_dotenv()\n",
    "\n",
    "required_keys = {\n",
    "    \"MODEL_PROVIDER\": os.getenv(\"MODEL_PROVIDER\"),\n",
    "    \"MODEL_NAME\": os.getenv(\"MODEL_NAME\"),\n",
    "    \"MODEL_BASE_URL\": os.getenv(\"MODEL_BASE_URL\"),\n",
    "    \"OLLAMA_API_KEY\": os.getenv(\"OLLAMA_API_KEY\"),\n",
    "    \"OPENWEATHERMAP_API_KEY\": os.getenv(\"OPENWEATHERMAP_API_KEY\"),\n",
    "    \"FOURSQUARE_API_KEY\": os.getenv(\"FOURSQUARE_API_KEY\"),\n",
    "    \"TAVILY_API_KEY\": os.getenv(\"TAVILY_API_KEY\"),\n",
    "    \"EXCHANGE_RATE_API_KEY\": os.getenv(\"EXCHANGE_RATE_API_KEY\"),\n",
    "}\n",
    "\n",
    "print(\"Project root:\", PROJECT_ROOT)\n",
    "print(\"\\nEnvironment key presence:\")\n",
    "for key, value in required_keys.items():\n",
    "    print(f\"{key:24} {'SET' if value else 'MISSING'}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a1ecbf2d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "API connectivity tests:\n",
      "\n",
      "PASS | OpenWeatherMap     | status=200, city=Goa\n",
      "FAIL | Foursquare         | status=401, results=0\n",
      "PASS | Tavily             | answer_length=537\n",
      "FAIL | ExchangeRate       | status=403, USD_INR=None\n",
      "PASS | LLM direct         | SmartTrip OK\n"
     ]
    }
   ],
   "source": [
    "import requests\n",
    "from langchain_tavily import TavilySearch\n",
    "from langchain_openai import ChatOpenAI\n",
    "from langchain_core.messages import HumanMessage\n",
    "\n",
    "\n",
    "def check_result(name, ok, detail):\n",
    "    status = \"PASS\" if ok else \"FAIL\"\n",
    "    print(f\"{status:4} | {name:18} | {detail}\")\n",
    "\n",
    "\n",
    "print(\"API connectivity tests:\\n\")\n",
    "\n",
    "# 1. OpenWeatherMap\n",
    "try:\n",
    "    weather_key = os.getenv(\"OPENWEATHERMAP_API_KEY\")\n",
    "    response = requests.get(\n",
    "        \"https://api.openweathermap.org/data/2.5/weather\",\n",
    "        params={\"q\": \"Goa\", \"appid\": weather_key, \"units\": \"metric\"},\n",
    "        timeout=20,\n",
    "    )\n",
    "    data = response.json()\n",
    "    check_result(\n",
    "        \"OpenWeatherMap\",\n",
    "        response.status_code == 200,\n",
    "        f\"status={response.status_code}, city={data.get('name', 'n/a')}\",\n",
    "    )\n",
    "except Exception as exc:\n",
    "    check_result(\"OpenWeatherMap\", False, repr(exc))\n",
    "\n",
    "# 2. Foursquare\n",
    "try:\n",
    "    fsq_key = os.getenv(\"FOURSQUARE_API_KEY\")\n",
    "    response = requests.get(\n",
    "        \"https://api.foursquare.com/v3/places/search\",\n",
    "        headers={\"Accept\": \"application/json\", \"Authorization\": fsq_key},\n",
    "        params={\"query\": \"tourist attraction\", \"near\": \"Goa\", \"limit\": 3},\n",
    "        timeout=20,\n",
    "    )\n",
    "    data = response.json()\n",
    "    result_count = len(data.get(\"results\", [])) if isinstance(data, dict) else 0\n",
    "    check_result(\n",
    "        \"Foursquare\",\n",
    "        response.status_code == 200 and result_count > 0,\n",
    "        f\"status={response.status_code}, results={result_count}\",\n",
    "    )\n",
    "except Exception as exc:\n",
    "    check_result(\"Foursquare\", False, repr(exc))\n",
    "\n",
    "# 3. Tavily\n",
    "try:\n",
    "    tavily = TavilySearch(topic=\"general\", include_answer=\"advanced\")\n",
    "    result = tavily.invoke({\"query\": \"top attractions in Goa India\"})\n",
    "    answer = result.get(\"answer\") if isinstance(result, dict) else str(result)\n",
    "    check_result(\"Tavily\", bool(answer), f\"answer_length={len(str(answer))}\")\n",
    "except Exception as exc:\n",
    "    check_result(\"Tavily\", False, repr(exc))\n",
    "\n",
    "# 4. ExchangeRate API\n",
    "try:\n",
    "    exchange_key = os.getenv(\"EXCHANGE_RATE_API_KEY\")\n",
    "    response = requests.get(\n",
    "        f\"https://v6.exchangerate-api.com/v6/{exchange_key}/latest/USD\",\n",
    "        timeout=20,\n",
    "    )\n",
    "    data = response.json()\n",
    "    rate = data.get(\"conversion_rates\", {}).get(\"INR\") if isinstance(data, dict) else None\n",
    "    check_result(\n",
    "        \"ExchangeRate\",\n",
    "        response.status_code == 200 and rate is not None,\n",
    "        f\"status={response.status_code}, USD_INR={rate}\",\n",
    "    )\n",
    "except Exception as exc:\n",
    "    check_result(\"ExchangeRate\", False, repr(exc))\n",
    "\n",
    "# 5. LLM direct call\n",
    "try:\n",
    "    llm = ChatOpenAI(\n",
    "        model=os.getenv(\"MODEL_NAME\"),\n",
    "        base_url=os.getenv(\"MODEL_BASE_URL\"),\n",
    "        api_key=os.getenv(\"OLLAMA_API_KEY\") or os.getenv(\"MODEL_API_KEY\"),\n",
    "    )\n",
    "    response = llm.invoke(\"Reply with exactly: SmartTrip OK\")\n",
    "    check_result(\"LLM direct\", bool(response.content), response.content[:120])\n",
    "except Exception as exc:\n",
    "    check_result(\"LLM direct\", False, repr(exc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2dd6a18b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Building graph...\n",
      "[ModelLoader] Loading LLM — provider: ollama\n",
      "Invoking full agent with tools...\n",
      "Agent response length: 11007\n",
      "\n",
      "Preview:\n",
      "\n",
      "## 5‑Day Goa Getaway for **2 people** – **₹ 10,000** total  \n",
      "*(≈ ₹ 2,000 per day – very budget‑friendly)*  \n",
      "\n",
      "---  \n",
      "\n",
      "### 1️⃣ QUICK OVERVIEW  \n",
      "\n",
      "| Item | Approx. Cost (INR) | Notes |\n",
      "|------|-------------------|-------|\n",
      "| **Accommodation** (5 nights, budget guesthouse/hostel) | **₹ 4,000** | ~₹ 800 per night for a double‑bed dorm or private room in Panaji/Anjuna |\n",
      "| **Food** (local shacks, street‑food, small restaurants) | **₹ 1,500** | ~₹ 150 per person per main meal (breakfast + dinner) + cheap snacks |\n",
      "| **Local transport** (state buses, shared autos, occasional scooter rent) | **₹ 800** | Bus fares ₹ 10‑₹ 30, scooter rent ₹ 250 / day for 2 days (optional) |\n",
      "| **Entry & activity fees** (temples, forts, waterfalls, market entry) | **₹ 800** | Mostly ₹ 20‑₹ 100 per site |\n",
      "| **Miscellaneous / Buffer** | **₹ 900** | Souvenirs, occasional water‑sport try‑out, emergency cash |\n",
      "| **Total** | **₹ 10,000** | ✔ Fits the budget |\n",
      "\n",
      "> **Tip:** All prices are 2026 averages; book early on sites like *Hostelworld* or *Airbnb* for the cheapest rooms, and pay for food & transport in cash to avoid extra card fees.\n",
      "\n",
      "---\n",
      "\n",
      "## 2️⃣ WEATHER (May 2026)\n",
      "\n",
      "| Parameter | Typical Range |\n",
      "|-----------|---------------|\n",
      "| **Temperature** | 30 °C – 34 °C (day) / 26 °C – 28 °C (night) |\n",
      "| **Rainfall** | Pre‑monsoon showers, ~30 mm total, usually brief afternoon thunders |\n",
      "| **What to pack** | Light cotton/linen clothes, a hat, sunscreen (SPF ≥ 30), a compact rain‑poncho, comfortable walking shoes or flip‑flops. |\n",
      "\n",
      "---\n",
      "\n",
      "## 3️⃣ ITINERARY – **Plan A (Main Tourist Highlights)**  \n",
      "\n",
      "| Day | Morning | Mid‑Day | Evening | Suggested Stay (Budget) |\n",
      "|-----|----------|--------|---------|--------------------------|\n",
      "| **1 – Arrival & Panaji** | Arrive at Dabolim Airport or bus‑station. Take a local bus to **Panaji** (₹ 20). | Check‑in, freshen up. Walk around **Fontainhas** (Portuguese‑style lanes). Lunch at **Café Mambo** – fish thali ₹ 150 pp. | Sunset at **Miramar Beach**; dinner at **Vinayak Family Restaurant** – Goan fish curry ₹ 120 pp. | **Hostel “Zostel Panaji”** – mixed dorm, ₹ 800/night (double). |\n",
      "| **2 – Old Goa & Fort Aguada** | Breakfast at **Ritz Café** – poha & tea ₹ 70 pp.<br>Bus to **Old Goa** (₹ 10). Visit **Basilica of Bom Jesus** (free) and **Se Cathedral** (free). | Lunch at **Martin’s Corner (Mapusa)** – chicken xacuti ₹ 180 pp. | Evening bus to **Fort Aguada** (₹ 20). Explore fort & lighthouse (free). Dinner at **Susegado** (rice & fish Curry) ₹ 120 pp. | Same hostel. |\n",
      "| **3 – No\n"
     ]
    }
   ],
   "source": [
    "from agent.agentic_workflow import GraphBuilder\n",
    "\n",
    "prompt = \"Plan a 5-day trip to goa for 2 people with a budget of inr 10000\"\n",
    "\n",
    "print(\"Building graph...\")\n",
    "graph = GraphBuilder(model_provider=os.getenv(\"MODEL_PROVIDER\", \"ollama\"))()\n",
    "\n",
    "print(\"Invoking full agent with tools...\")\n",
    "output = graph.invoke({\"messages\": [HumanMessage(content=prompt)]})\n",
    "answer = output[\"messages\"][-1].content\n",
    "\n",
    "print(\"Agent response length:\", len(answer))\n",
    "print(\"\\nPreview:\\n\")\n",
    "print(answer[:2500])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "trip_env (3.12.9)",
   "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": "3.12.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}