🌐 Google Trends MCP Server

The Real-Time Google Trends Model Context Protocol (MCP) Server for AI Assistants.

Connect Claude Desktop, Cursor, Antigravity, Windsurf, and Cline directly to live global search trends, breaking news, publisher sources, and extracted article summaries.

Features • Installation • Client Setup • Tools • Resources • Prompts • Docker

⚡ Highlights & Features

• ⚡ Zero-Config & Instant Boot (~150-300ms): No database (PostgreSQL) or Redis cache setup required. Runs out-of-the-box directly querying Google Trends real-time streams.

• 📰 Automated Article Summaries & Sources: Fetches not only trend queries, but also news headlines, publisher/source names, direct source links, and automatically extracted article summaries.

• 📖 Deep Article Reader: Powered by Cheerio and Playwright with Ghostery ad-blocking to read, clean, and convert any news article into clean text/markdown for the LLM.

• 🌍 Global Coverage (~250 Countries): Query trends across all ISO 3166-1 alpha-2 country codes (TR, US, GB, DE, FR, JP, CA, etc.).

• 🏷️ Category Filtering: Filter by all, business, tech (Sci/Tech), sports, health, entertainment, and top stories.

• 🔌 Dual Transports:

  • stdio (Default): Standard input/output for local desktop clients (Claude Desktop, Cursor, Antigravity, Cline).

  • SSE (Server-Sent Events): HTTP transport on port 3333 for remote microservices and distributed agents.

Related MCP server: news-sentiment-mcp

🏗️ Architecture

┌────────────────────────────────────────────────────────────────────────┐
│             AI Assistants (Claude Desktop, Cursor, Antigravity, Cline) │
└───────────────────────────────────┬────────────────────────────────────┘
                                    │  JSON-RPC 2.0 (stdio / SSE)
                                    ▼
┌────────────────────────────────────────────────────────────────────────┐
│                   Google Trends MCP Server (bin/mcp-server.js)         │
│          • Tools (Trends, Article Reader, Metadata)                    │
│          • Resources (trends://latest, trends://countries)             │
│          • Prompts (analyze_trends, trend_deep_dive)                   │
└───────────────────────────────────┬────────────────────────────────────┘
                                    │
                                    ▼
┌────────────────────────────────────────────────────────────────────────┐
│                   Real-Time Fetch & Article Engine                     │
│   • Live Google Trends RSS / JSON Stream Parser (~200ms)               │
│   • Concurrent HTML Metadata & Summary Extractor                       │
│   • Headless Chromium (Playwright + AdBlocker) Web Reader Fallback     │
└────────────────────────────────────────────────────────────────────────┘

🚀 Quick Start

1. Direct Run with npx / node

You can run the MCP server immediately without installing any external services:

# Clone the repository
git clone https://github.com/yusuf-polat/Google-Trends.git
cd Google-Trends

# Install dependencies
npm install

# Start in stdio mode (Default for AI Desktop clients)
npm start

# Or start in HTTP SSE mode
npm run mcp:sse

2. Browser Testing via MCP Inspector

Test all tools interactively in your browser with the official MCP Inspector:

npx @modelcontextprotocol/inspector node bin/mcp-server.js

🔌 AI Client Setup Guides

1. Claude Desktop

Add the server configuration to your claude_desktop_config.json:

• Windows: %APPDATA%\Claude\claude_desktop_config.json

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "google-trends": {
      "command": "node",
      "args": ["<PATH-TO-REPO>/bin/mcp-server.js"]
    }
  }
}

(Replace <PATH-TO-REPO> with your absolute path, e.g. D:/Google-Trends)

2. Cursor IDE

Create or edit .cursor/mcp.json in your project or global directory:

{
  "mcpServers": {
    "google-trends": {
      "command": "node",
      "args": ["<PATH-TO-REPO>/bin/mcp-server.js"]
    }
  }
}

Or go to Settings > Features > MCP > Add New MCP Server and select command.

3. Antigravity IDE

Add to your global ~/.gemini/config/mcp_config.json:

{
  "mcpServers": {
    "google-trends": {
      "command": "node",
      "args": ["<PATH-TO-REPO>/bin/mcp-server.js"]
    }
  }
}

4. VS Code (Cline / Roo Code / Continue)

In your Cline / Roo Code extension settings (cline_mcp_settings.json):

{
  "mcpServers": {
    "google-trends": {
      "command": "node",
      "args": ["<PATH-TO-REPO>/bin/mcp-server.js"]
    }
  }
}

🛠️ Available MCP Tools

1. get_realtime_trends

Fetch live trending topics and searches with approximate search volume, breakdown queries, publisher sources, and extracted article summaries.

Sample Response:

{
  "source": "google_trends_realtime_rss",
  "country": "TR",
  "country_name": "Türkiye (Turkey)",
  "returned_count": 1,
  "trends": [
    {
      "trend_name": "sıla türkoğlu",
      "search_volume": "5000+",
      "published_at": "Thu, 17 Sep 2026 07:20:00 -0700",
      "news": [
        {
          "title": "650 bin TL'lik davada mahkeme kararını verdi",
          "publisher": "Sözcü Gazetesi",
          "url": "https://www.sozcu.com.tr/...",
          "summary": "Sıla Türkoğlu'nun Kanal 7'de yayınlanan 'Emanet' dizisinden ayrılmasının ardından açılan davada mahkeme kararını verdi...",
          "cover_image": "https://sozcu01.sozcucdn.com/..."
        }
      ]
    }
  ]
}

2. get_trend_details

Get an in-depth breakdown of a single trending topic, inspecting all of its associated news stories with full summaries and publishers.

3. extract_article_content

Extract clean full-text content, headline, summary, and metadata from any news URL linked to a trend. Automatically bypasses cookie popups and ad blocks.

4. list_supported_countries

List all ~250 ISO 3166-1 alpha-2 country codes and country names supported by Google Trends.

5. list_supported_categories

List supported Google Trends category shortcuts (business, entertainment, health, tech, sports, top).

6. get_trends_stats

Returns system operational metrics and engine capability summary.

📦 MCP Resources

Clients can read live resources directly without invoking tool arguments:

💡 MCP Prompts

Pre-engineered prompt templates for AI workflows:

• analyze_trends: Guides the AI to fetch current trends, cluster them into thematic domains (Politics, Tech, Markets, Entertainment), and produce an executive briefing.

• trend_deep_dive: Instructs the AI to perform investigative reporting on a specific trend, reading linked articles and identifying stakeholders, timelines, and perspectives.

🐳 Docker Deployment

To run Google Trends MCP as an isolated container:

# Build Docker image
docker build -t google-trends-mcp .

# Run in SSE HTTP mode (Port 3333)
docker run -d -p 3333:3333 --name google-trends-mcp google-trends-mcp npm run mcp:sse

Once running, configure remote MCP clients using the SSE URL: http://localhost:3333/sse

💬 Example Prompt Interactions

Ask your AI assistant naturally:

User: "What are the top 5 trending topics in Turkey right now, and what are their search volumes?" (Assistant calls get_realtime_trends with country: "TR", limit: 5)

User: "What's trending in Tech in the United States today? Summarize the main article for the #1 topic." (Assistant calls get_realtime_trends with country: "US", category: "tech" and summarizes the news)

User: "Read this news article from the trend and give me the full breakdown: https://..." (Assistant calls extract_article_content)

📄 License

This project is licensed under the ISC License. Made with ❤️ for the open-source MCP and AI community.