AI Weather Agent with MCP, Gemini and Playwright

Overview

This project implements an AI Agent that retrieves current weather information for cities in Israel using the Model Context Protocol (MCP), Google Gemini, and Playwright.

When the user asks for the weather in a city, the agent opens a weather website, searches for the requested city, extracts the weather information from the page, and returns an answer based on the website's content.

The project also includes a simple retrieval step, where the weather information is extracted from the web page and provided to the language model as additional context before generating the final response.

Technologies

  • Python
  • Google Gemini API
  • MCP (Model Context Protocol)
  • Playwright
  • AsyncIO
  • python-dotenv

Project Structure

host.py                 # Main chat application
client.py               # MCP client
weather_Israel.py       # MCP weather server
README.md
requirements.txt

How It Works

  1. The user asks for the weather in an Israeli city.
  2. Gemini selects the appropriate MCP tools.
  3. Playwright opens the weather website.
  4. The requested city is entered into the search box.
  5. The city is selected from the autocomplete list.
  6. The weather page is loaded.
  7. The weather information is extracted from the page.
  8. The extracted text is returned to Gemini.
  9. Gemini generates the final answer using the extracted information.

Features

  • Supports weather queries for cities in Israel.
  • Accepts city names in Hebrew or English.
  • Uses Playwright to navigate a live weather website.
  • Extracts weather information directly from the website.
  • Provides the extracted content to the language model before answering.
  • Uses the website content instead of relying on the model's internal knowledge.

Installation

Clone the repository:

git clone https://github.com/YisNiz/AI-MCP-with-RAG-project

Install the required packages:

pip install -r requirements.txt

Create a .env file containing:

GEMINI_API_KEY=YOUR_API_KEY

Install Playwright browsers:

playwright install

Run the project:

python host.py

Example Questions

  • What is the weather in Jerusalem?
  • Weather in Tel Aviv
  • מזג האוויר בחיפה
  • מה מזג האוויר בבאר שבע?

Notes

  • The weather information is retrieved from a live weather website.
  • The language model is instructed to answer using the extracted website content.
  • The project demonstrates the integration of an LLM with MCP tools and browser automation.

Author

Yisca Nazri

MCP Server · Populars

MCP Server · New

    mobbin

    Official Mobbin MCP server

    Official Mobbin MCP server repository

    Community mobbin
    frankchu91

    MindBase — Karpathy's LLM Wiki, as a product

    Karpathy's LLM Wiki idea as a product — an AI that builds and maintains a markdown wiki from your notes and sources. MCP server + web UI, runs on free local models (Ollama), no API key needed. MIT.

    Community frankchu91
    aakarim

    📜 OpenLore

    A minimal, extensible, agent-native knowledge base that keeps shared context current and inspectable

    Community aakarim
    sv-grid

    @svgrid/mcp

    Native Svelte 5 data grid. Headless-first engine + drop-in render component. Row + column virtualization (1M rows), Excel-style filters, inline editing, grouping, pivot, server-side data. MIT core (@svgrid/grid), MCP server for Claude / Cursor. https://svgrid.com

    Community sv-grid
    cinderline

    NorthCinder

    Buyer-run, ad-neutral shopping-agent MCP software with deterministic ranking, signed purchase mandates, and a local audit trail.

    Community cinderline