# 🌍 Global News MCP Server

![Python](https://img.shields.io/badge/Python-3.11-blue)
![MCP](https://img.shields.io/badge/Model_Context_Protocol-MCP-green)
![Streamlit](https://img.shields.io/badge/UI-Streamlit-red)
![RSS](https://img.shields.io/badge/Data-RSS_Feeds-orange)

A **Model Context Protocol (MCP) powered global news application** that collects news from multiple countries and categories using RSS feeds and exposes them as reusable MCP tools.

The project contains:

- 🧠 MCP Server for news intelligence
- 🌐 Streamlit web interface
- 📰 RSS-based news aggregation
- 🔎 Global news search
- 🌍 Country-based filtering
- 💻 Technology and science news categories

No API keys are required.

---

# ✨ Features

## 🌎 Global News

Fetch worldwide news from trusted RSS sources:

- BBC World
- The Guardian


## 🌍 Country News

Get news by country:

| Country | Sources |
|---|---|
| 🇮🇳 India | The Hindu, Times of India |
| 🇺🇸 USA | CNN |
| 🇬🇧 UK | BBC UK |
| 🇨🇦 Canada | CBC |
| 🇦🇺 Australia | ABC Australia |
| 🇩🇪 Germany | DW |
| 🇫🇷 France | France24 |


## 💻 Technology News

Technology updates from:

- TechCrunch
- The Verge


## 🚀 Science News

Science and space updates from:

- NASA
- Science Daily


## 🔎 Global Search

Search news articles across multiple RSS feeds.

Examples:

```
Artificial Intelligence
Space
Climate Change
Robotics
Technology
```

---

# 🏗️ Architecture

```
                    User
                     |
                     |
              Streamlit UI
                     |
                     |
                MCP Client
                     |
                     |
                MCP Server
                     |
          -----------------------
          |                     |
    News Service            MCP Tools
          |
          |
     RSS Feed Sources
          |
          |
    Structured JSON Response
```

---

# 🔌 MCP Tools

The MCP Server exposes these tools:

---

## `get_global_news()`

Returns latest worldwide news.

Example response:

```json
[
 {
  "source": "BBC World",
  "title": "Latest World News",
  "summary": "News summary",
  "link": "article url"
 }
]
```

---

## `get_country_news(country)`

Fetch news from a specific country.

Example:

```
get_country_news("India")
```

Returns:

```json
[
 {
  "source": "The Hindu",
  "title": "India News",
  "summary": "...",
  "link": "..."
 }
]
```

---

## `get_technology_news()`

Returns technology-related articles.

Sources:

- TechCrunch
- The Verge

---

## `get_science_news()`

Returns science-related articles.

Sources:

- NASA
- Science Daily

---

## `search_news(keyword)`

Searches across available RSS feeds.

Example:

```
search_news("AI")
```

---

# 📂 Project Structure

```
news-mcp/

│
├── app.py                    # Streamlit frontend
│
├── requirements.txt          # Dependencies
│
├── mcp_server/
│   │
│   ├── __init__.py
│   ├── server.py             # MCP server and tools
│   └── news_service.py       # RSS processing logic
│
└── components/
    │
    ├── __init__.py
    └── cards.py              # News UI components
```

---

# 🛠️ Tech Stack

## Backend

- Python 3.11
- Model Context Protocol (MCP)
- FastMCP
- Feedparser


## Frontend

- Streamlit


## Data Sources

RSS feeds from:

- BBC
- The Guardian
- The Hindu
- CNN
- CBC
- DW
- France24
- TechCrunch
- NASA
- Science Daily

---

# ⚙️ Installation

## Clone Repository

```bash
git clone https://github.com/kevin-arock/news-mcp.git
```

Navigate:

```bash
cd news-mcp
```

---

## Create Virtual Environment

```bash
python -m venv venv
```

Activate environment.

### Windows

```bash
venv\Scripts\activate
```

### Linux / Mac

```bash
source venv/bin/activate
```

---

## Install Dependencies

```bash
pip install -r requirements.txt
```

---

# ▶️ Run Application

Start Streamlit:

```bash
streamlit run app.py
```

Application opens at:

```
http://localhost:8501
```

---

# 🔄 Application Flow

```
User selects news category
            |
            |
      Streamlit Interface
            |
            |
        MCP Client
            |
            |
        MCP Server
            |
            |
       News MCP Tools
            |
            |
       RSS News Sources
            |
            |
      JSON News Response
            |
            |
        UI Cards
```

---

# 💡 Why MCP?

Instead of directly connecting the UI to data sources, MCP creates a reusable tool layer.

Advantages:

✅ AI agents can discover and call tools  
✅ Backend capabilities are separated from UI  
✅ Multiple AI clients can use the same tools  
✅ External data sources can be integrated easily  

---

# 🎯 Interview Overview

> "I built a Global News MCP Server that exposes news retrieval capabilities as MCP tools. The server collects RSS feeds from different countries, processes them into structured JSON responses, and allows a Streamlit MCP client to consume those tools. This demonstrates how AI applications can use MCP as a standard way to connect with external capabilities."

---

# 🚀 Future Improvements

Planned enhancements:

- 🤖 AI-powered article summarization
- 🌐 Multi-language translation
- 🧠 Semantic search using embeddings
- 📊 News sentiment analysis
- 🔔 Personalized news recommendations
- 💬 Chat-based AI news assistant
- 🔗 Integration with AI clients supporting MCP

---

# 👨‍💻 Author

## Arockia Kevin J

GitHub:

https://github.com/kevin-arock


---

# ⭐ Support

If you like this project, consider giving it a ⭐ on GitHub.