tradingview

This MCP server provides real-time market intelligence and technical analysis by leveraging TradingView data. It supports a range of tools for market screening, technical analysis, and information lookups across multiple exchanges and timeframes. You can query top gainers/losers, run Bollinger Band and RSI based scans, and perform in-depth technical analyses on specified assets. The server is designed to be controlled via natural language prompts or structured tool calls, enabling integration with AI assistants like Claude or other MCP clients.

To use it, deploy the server through UV as shown in the installation guide, then send requests that reference the available tools (e.g., top_gainers, bollinger_scan, coin_analysis, rating_filter, or exchanges://list). Examples include asking for the top crypto gainers on KuCoin in the last 15 minutes, analyzing BTC with multiple indicators, or scanning for assets with a Bollinger Band squeeze. The system supports multi-timeframe data (5m to 1M) and cross-market insights across cryptocurrencies and traditional markets (e.g., NASDAQ, BIST).

Notes: results may be subject to exchange data availability and API rate limits. If you encounter limited data or rate limit messages, try different exchanges, timeframes, or symbols, and respect any throttling recommendations provided by the server.

Prerequisites:

• Git
• Python (for the TradingView MCP) and UV package manager (as used in the examples)
• Internet access to clone the repository and install dependencies

Option A: UV-based manual setup (recommended for Claude Desktop integration)

1. Clone the repository:

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

2. Install UV (if not already installed) and sync dependencies:

# Install UV (example for direct installation as shown in README)
# macOS/Linux:
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows users can follow the Windows-specific instructions in INSTALLATION.md
uv sync

3. Run the MCP server locally (example invocation):

uv run python src/tradingview_mcp/server.py

4. Optional: configure Claude Desktop to connect via UV as shown in the README examples under Option 2 and local configuration blocks.

Option B: Claude Desktop integration (from README)

1. Install UV on your system (as above).
2. Add the server to Claude Desktop configuration using one of the JSON examples provided in the README (tradingview-mcp-local or tradingview-mcp for remote access).
3. Restart Claude Desktop to load the new MCP server.

Prerequisites recap:

• Git to clone the repository
• Python environment compatible with the server
• UV package manager to run the MCP server (uv sync and uv run)
• Network access to fetch dependencies and communicate with Claude or other MCP clients

Tips and caveats:

• If you encounter 'No data found' or empty arrays, try different supported exchanges (KuCoin often has more reliable data) and standard timeframes (15m, 1h, 1D).
• Rate limits can affect responses. If you hit limits, wait a few minutes between queries and let the server retry automatically.
• Claude Desktop discovery may require you to restart the app after adding the MCP configuration.
• Ensure your cwd (working directory) is correctly set when configuring the server in Claude Desktop or your MCP client.
• The available tools include market screening (top_gainers, top_losers, bollinger_scan, rating_filter), technical analysis (coin_analysis, consecutive_candles_scan, advanced_candle_pattern), and information (exchanges://list).
• Timeframe support includes 5m, 15m, 1h, 4h, 1D, 1W, 1M; adjust your queries accordingly for best results.