ali-rajabpour/tradingview-mcp

TradingView MCP Server 📈

A lightweight Model Context Protocol (MCP) server for fetching TradingView chart snapshots. Uses Playwright for efficient browser automation with persistent sessions and cookie-based authentication.

Note: This uses Playwright because TradingView renders charts client-side with JavaScript/Canvas. There is no pure HTTP API for chart images. This is the lightest possible working solution (~150MB RAM vs ~500MB with Selenium).

✨ Features

• 🪶 Lightweight: Playwright in headless mode (~150MB vs ~500MB with Selenium)
• 🚀 Fast: Persistent browser reuse, 3-5 seconds per chart
• 🔐 Secure: Session-based authentication via cookies
• 🎨 Customizable: Configure chart dimensions, intervals, and themes
• 🔧 MCP Compatible: Works with any MCP-enabled client (Claude Desktop, etc.)
• ♻️ Efficient: Reuses browser instances across multiple requests

📋 Prerequisites

• Python 3.10 or higher
• A TradingView account (free or paid)
• TradingView session cookies

Table of Contents

• Installation
• Quick Start
• Usage
• Configuration
• Available Tools
• Symbol Format
• Troubleshooting
• Performance
• Contributing

🚀 Installation

1. Clone the repository

git clone https://github.com/yourusername/tradingview-mcp.git
cd tradingview-mcp

2. Create a virtual environment

# Windows
python -m venv venv
.\venv\Scripts\activate

# Linux/Mac
python3 -m venv venv
source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

# Install Playwright browsers (one-time setup, ~150MB download)
python -m playwright install chromium

4. Configure credentials

1. Copy .env.example to .env:

   # Windows
   copy .env.example .env
   
   # Linux/Mac
   cp .env.example .env
   

2. Get your TradingView session cookies:

   • Log into TradingView in your browser
   • Press F12 to open Developer Tools
   • Go to Application tab (Chrome) or Storage tab (Firefox)
   • Navigate to Cookies > https://www.tradingview.com
   • Find and copy these two cookie values:
     • sessionid → Copy the entire value
     • sessionid_sign → Copy the entire value

   ⚠️ Important: Copy the full values including any special characters (slashes, equals signs, etc.). Don't add quotes or extra spaces.

3. Edit .env and paste your values:

   TRADINGVIEW_SESSION_ID=your_actual_session_id_here
   TRADINGVIEW_SESSION_ID_SIGN=your_actual_session_id_sign_here
   

   Example (with fake values):

   TRADINGVIEW_SESSION_ID=47zlhkbl2weohrhmjgufeg1o24droaod
   TRADINGVIEW_SESSION_ID_SIGN=v3:6S2OD9ta349/yqN0RdCeK3UKxTl/tJr4AOoPRUa0Crk=
   

5. Test your setup (Optional but recommended)

Create a test file test.py:

import asyncio
import sys
from pathlib import Path
sys.path.insert(0, str(Path(__file__).parent / "src"))

from dotenv import load_dotenv
load_dotenv()

from tradingview_mcp.server import get_chart_snapshot

async def test():
    image = await get_chart_snapshot("BINANCE:BTCUSDT", "D", 1200, 600, "dark")
    if image:
        Path("test_chart.png").write_bytes(image)
        print(f"✅ Success! Chart saved to test_chart.png ({len(image)} bytes)")
    else:
        print("❌ Failed to fetch chart")

asyncio.run(test())

Run it:

python test.py

If successful, you'll see test_chart.png with a Bitcoin chart!

🎯 Usage

Quick Start - Testing Locally

After installation, test the server:

python src/tradingview_mcp/server.py

The server will start and wait for MCP client connections via stdio.

Configuration for Claude Desktop

Add this to your Claude Desktop config file:

Windows: %APPDATA%\Claude\claude_desktop_config.json
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Linux: ~/.config/Claude/claude_desktop_config.json

Option 1: Using .env file (Recommended)

{
  "mcpServers": {
    "tradingview": {
      "command": "python",
      "args": ["C:\\path\\to\\tradingview-mcp\\src\\tradingview_mcp\\server.py"],
      "cwd": "C:\\path\\to\\tradingview-mcp"
    }
  }
}

Option 2: Inline credentials

{
  "mcpServers": {
    "tradingview": {
      "command": "python",
      "args": ["C:\\path\\to\\tradingview-mcp\\src\\tradingview_mcp\\server.py"],
      "env": {
        "TRADINGVIEW_SESSION_ID": "your_session_id",
        "TRADINGVIEW_SESSION_ID_SIGN": "your_session_id_sign"
      }
    }
  }
}

💡 Tip: Use absolute paths. On Windows, use double backslashes \\ or forward slashes /.

Restart Claude Desktop and you'll see the TradingView tools available!

Example Queries in Claude

Once configured, you can ask Claude:

- "Get me a daily chart of Bitcoin"
- "Show me AAPL on 1-hour timeframe"
- "Fetch a weekly chart for NASDAQ:TSLA with light theme"
- "Get BINANCE:ETHUSDT 5-minute chart"
- "Validate my TradingView session"
- "What timeframes are available?"

🛠️ Available Tools

1. get_chart_snapshot

Fetch a TradingView chart snapshot.

Parameters:

• symbol (required): Trading pair in TradingView format
  • Examples: "BINANCE:BTCUSDT", "NASDAQ:AAPL", "BITSTAMP:BTCUSD"
• interval (optional): Chart timeframe (default: "D")
  • Minutes: "1", "5", "15", "30", "60", "240"
  • Days/Weeks/Months: "D", "W", "M"
• width (optional): Image width in pixels (default: 1200)
• height (optional): Image height in pixels (default: 600)
• theme (optional): "dark" or "light" (default: "dark")

Example:

Get me a daily chart of Bitcoin on Binance

2. validate_session

Check if your TradingView session credentials are valid.

Example:

Validate my TradingView session

3. list_timeframes

List all available chart timeframes/intervals.

Example:

What timeframes are available?

📊 Symbol Format

TradingView uses the format: EXCHANGE:SYMBOL

Common Examples

Finding Symbol Names

1. Go to TradingView
2. Search for your desired asset
3. Look at the URL or chart title for the format EXCHANGE:SYMBOL

🔧 Architecture

This MCP server uses Playwright for lightweight browser automation:

1. Authentication: Session cookies passed to browser context
2. Chart Generation: Playwright navigates to TradingView and captures screenshots
3. Efficiency: Persistent browser instances reused across requests
4. MCP Protocol: Communicates via stdio with MCP clients

Why Playwright (Not Pure HTTP)?

• TradingView renders charts client-side with JavaScript/Canvas
• No server-side API exists for chart image generation
• Playwright is lighter than Selenium (~150MB vs ~500MB RAM)
• Headless mode with optimized flags minimizes resource usage
• Browser reuse makes subsequent requests very fast (1-2 seconds)

✅ This is the lightest possible approach for TradingView charts

🐛 Troubleshooting

Cookie/Authentication Issues

Problem: "Session credentials are not working" or "Invalid session"

Solutions:

1. Check for trailing commas or spaces: Make sure your .env file has no extra characters

   # ❌ Wrong (has comma)
   TRADINGVIEW_SESSION_ID=abc123,
   
   # ✅ Correct
   TRADINGVIEW_SESSION_ID=abc123
   

2. Refresh cookies: Session cookies expire after ~30 days

   • Log out and back into TradingView
   • Get fresh cookies from Developer Tools
   • Update your .env file

3. Check you're logged in: Make sure you're logged into TradingView when copying cookies

4. Copy entire value: Include special characters (=, /, :, etc.)

Symbol Not Found

Problem: "Failed to fetch chart snapshot"

Solutions:

1. Use correct format: EXCHANGE:SYMBOL (all uppercase)

   • ✅ BINANCE:BTCUSDT
   • ❌ btcusdt or binance:btcusdt

2. Verify symbol exists on TradingView.com

3. Test with a known symbol first: BINANCE:BTCUSDT or NASDAQ:AAPL

Timeout Errors

Problem: "Timeout exceeded" or slow performance

Solutions:

1. Check internet connection and TradingView.com accessibility
2. First request takes longer (5-8s) while browser starts
3. Subsequent requests are faster (3-5s) due to browser reuse
4. If persistent, restart the MCP server

Playwright Installation Issues

Problem: "Playwright not found" or browser download fails

Solutions:

# Reinstall Playwright
pip uninstall playwright
pip install playwright

# Install browsers again
python -m playwright install chromium

# If still failing, install system dependencies (Linux)
sudo playwright install-deps

Claude Desktop Not Recognizing Server

Problem: Tools don't appear in Claude Desktop

Solutions:

1. Check JSON syntax in config file (use a JSON validator)
2. Use absolute paths (not relative)
3. Verify Python path is correct:

   # Windows
   where python
   
   # Mac/Linux
   which python
   

4. Completely restart Claude Desktop (not just reload)
5. Check Claude logs for errors:
   • Windows: %APPDATA%\Claude\logs\
   • Mac: ~/Library/Logs/Claude/

Performance Issues

Problem: High memory usage or slow responses

Tips:

• Memory usage: ~150MB is normal for Playwright
• First chart: 5-8 seconds (browser startup)
• Subsequent charts: 3-5 seconds
• Browser is reused across requests for efficiency
• If memory grows over time, restart the server

⚡ Performance

Comparison with Alternatives:

🔒 Security Notes

• ⚠️ Never commit your .env file - it contains your session credentials
• ⚠️ Don't share your session cookies - they provide full account access
• ⚠️ Rotate cookies regularly - they expire after ~30 days
• ✅ .env is in .gitignore by default
• ✅ Use environment variables or secure credential management in production

📝 Development

Project Structure

tradingview-mcp/
├── src/
│   └── tradingview_mcp/
│       ├── __init__.py
│       └── server.py          # Main MCP server
├── .env.example               # Template for credentials
├── .gitignore
├── README.md
├── requirements.txt           # Python dependencies
└── pyproject.toml            # Project metadata

Running Tests

# Install dev dependencies
pip install -e ".[dev]"

# Run tests
pytest

Code Formatting

# Format code
black src/

# Lint code
ruff check src/

🤝 Contributing

Contributions are welcome! Here's how:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes
4. Test thoroughly
5. Commit your changes (git commit -m 'Add amazing feature')
6. Push to the branch (git push origin feature/amazing-feature)
7. Open a Pull Request

Development Setup:

pip install -e ".[dev]"
black src/
ruff check src/

📄 License

MIT License - see file for details.

🙏 Acknowledgments

• TradingView - Excellent charting platform
• Anthropic - Model Context Protocol
• Playwright - Browser automation

⚠️ Disclaimer

This is an unofficial tool and is not affiliated with, endorsed by, or connected to TradingView. Use at your own risk and in accordance with TradingView's Terms of Service. Respect rate limits and be mindful of your usage.

� Support & Issues

• 🐛 Bug Reports: GitHub Issues
• 💬 Questions: GitHub Discussions
• 📧 Email: your.email@example.com

📊 Project Stats

• Version: 0.1.0
• Python: 3.10+
• Memory: ~150MB
• Response Time: 3-5 seconds
• License: MIT

Made with ❤️ by the trading community

Star ⭐ this repo if you find it useful!