TradingView MCP Server

Content

TradingView MCP Server exposes market-research tools to Claude and other MCP clients through the tradingview-mcp Python entry point. It can scan crypto and stock markets, run Bollinger Band and candle-pattern analysis, fetch Yahoo Finance prices, summarize financial news, inspect Reddit sentiment, compare backtest strategies, run walk-forward checks, and retrieve options-chain or unusual-options activity for supported symbols.

The project is designed for analysis and education, not trade execution. Its README includes an explicit financial-advice disclaimer and positions the server as an informational tool whose outputs must be independently verified before any financial decision.

Source Review

• https://github.com/atilaahmettaner/tradingview-mcp
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/README.md
• https://pypi.org/pypi/tradingview-mcp-server/json
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/INSTALLATION.md
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/pyproject.toml
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/.codex-mcp.json
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/src/tradingview_mcp/server.py
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/src/tradingview_mcp/core/services/backtest_service.py
• https://github.com/atilaahmettaner/tradingview-mcp/blob/main/LICENSE
• https://cryptosieve.com/mcp-server

These sources were reviewed on 2026-06-06. Prefer the live repository, README, installation guide, PyPI package metadata, Codex MCP config, Python package metadata, MCP server implementation, backtesting service, license file, and project homepage for current install commands, tool coverage, hosted-option context, financial disclaimers, and licensing.

Features

• Python package tradingview-mcp-server with the tradingview-mcp entry point.
• MCP configuration for uvx --from tradingview-mcp-server tradingview-mcp.
• Market screeners for top gainers, top losers, Bollinger squeezes, ratings, volume breakouts, candle patterns, and multi-timeframe analysis.
• Technical-analysis workflows for crypto exchanges and stock markets.
• Yahoo Finance price and market snapshot helpers.
• Financial news, Reddit sentiment, and combined analysis tools.
• Strategy backtesting, strategy comparison, walk-forward validation, trade-log, and equity-curve workflows.
• Options-chain and unusual-options activity tools for US stock symbols.
• MIT license.

Installation

Install or launch the MCP package with uv:

uvx --from tradingview-mcp-server tradingview-mcp

Configure an MCP client:

{
  "mcpServers": {
    "tradingview": {
      "command": "uvx",
      "args": ["--from", "tradingview-mcp-server", "tradingview-mcp"]
    }
  }
}

The upstream documentation also includes source-based and Windows-specific setup paths. Use the tradingview-mcp-server package name; tradingview-mcp on PyPI is a different package from a different repository.

Use Cases

• Ask Claude for a top-gainers or top-losers scan for a supported exchange.
• Run Bollinger Band, candle-pattern, volume-breakout, or multi-timeframe analysis for a symbol.
• Compare technical-analysis output with Reddit sentiment and current financial news for research.
• Backtest a supported strategy and compare the result with a walk-forward validation check.
• Inspect Yahoo Finance prices, broad market snapshots, extended-hours moves, options chains, or unusual options activity.
• Explore crypto and stock market data without wiring Claude to a brokerage or trade-execution API.

Safety and Privacy

TradingView MCP Server should be treated as an analysis assistant, not a trading system. Its outputs can contain trade-plan language, ratings, indicators, targets, stop-loss-like values, sentiment summaries, and backtest statistics, but those outputs are not financial advice and can be wrong. Independently verify data quality, timestamp freshness, exchange coverage, assumptions, fees, slippage, and provider terms before using any result.

Keep execution separate from analysis. Do not connect MCP-generated conclusions directly to automated order placement without a separate human-approved trading, risk, and compliance process. Watch for rate limits, provider outages, delayed feeds, symbol mismatches, stale prices, and overfit backtests.

Prompts and results can reveal watched symbols, market interests, strategy ideas, timeframes, private research notes, and position-like intent. Do not paste brokerage credentials, balances, order history, account identifiers, API keys, or confidential portfolio information into the MCP client. The self-hosted server runs locally, but its tools still call external market, news, and sentiment data sources; the upstream hosted connector is a separate commercial option with its own trust boundary.

Duplicate Check

Existing MCP content includes finance and market-data entries such as Financial Datasets MCP Server, SEC EDGAR entries, Drillr, and blockchain-oriented servers. TradingView MCP Server is distinct because it documents atilaahmettaner/tradingview-mcp and the tradingview-mcp-server package for TradingView-style screeners, technical indicators, sentiment/news aggregation, backtesting, walk-forward validation, Yahoo Finance lookups, and options-chain research. No dedicated TradingView MCP entry was found in content/mcp.