metatrader-mcp-server
Enables natural language management of MetaTrader 5 accounts, including checking balances, opening positions, and managing trades via the Antigravity agent.
README
UT Bot MetaTrader 5 Webhook Bridge & AI Agent
Automate your TradingView Pine Script alerts directly into MetaTrader 5 (MT5) with custom Stop Loss (SL) and Take Profit (TP) levels using FastAPI, ngrok, and the Antigravity developer agent.
🌟 How It Works
[TradingView Strategy Alert]
│ (Triggers & sends JSON payload)
▼
[ngrok Tunnel (Public URL)]
│ (Forwards to local machine)
▼
[FastAPI Webhook Server]
│ (Parses payload & gets current price)
▼
[MetaTrader 5 Client]
│ (Executes market order with SL & TP)
▼
[MT5 Terminal]
📋 Prerequisites
- Windows OS (required for the MetaTrader 5 Python SDK).
- Python 3.10+ installed.
- MetaTrader 5 Desktop Terminal installed and logged into your broker account.
- Algorithmic Trading Enabled:
- Open MT5.
- Go to Tools ➔ Options ➔ Expert Advisors.
- Check "Allow algorithmic trading" and click OK.
🚀 Quick Start & Installation
Step 1: Install Dependencies
Install the required Python packages:
pip install fastapi uvicorn metatrader-mcp-server pydantic
Step 2: Configure Antigravity MCP Server
To allow Antigravity to check your balances, open positions, and manage trades using natural language, update your global Antigravity configuration file:
File Path: C:\Users\DATA ENG. OLA\.gemini\config\mcp_config.json
Add the "metatrader" server configuration:
{
"mcpServers": {
"metatrader": {
"command": "C:\\Users\\DATA ENG. OLA\\AppData\\Local\\Packages\\PythonSoftwareFoundation.Python.3.13_qbz5n2kfra8p0\\LocalCache\\local-packages\\Python313\\Scripts\\metatrader-mcp-server.exe",
"args": [
"--login", "YOUR_MT5_LOGIN",
"--password", "YOUR_MT5_PASSWORD",
"--server", "YOUR_MT5_SERVER",
"--transport", "stdio"
]
}
}
}
(Make sure to adjust the path to your python local-packages directory if it differs).
Step 3: Set Up the Webhook Bridge
Create the webhook_bridge.py file to receive and process alerts from TradingView. The script automatically handles symbol cleaning (removing exchange prefixes like EXNESS:) and resolves library-specific SL/TP validation bugs.
Running the Server
Run the webhook bridge on port 5001:
python webhook_bridge.py
Step 4: Expose the Port Using ngrok
Expose port 5001 to the internet to get a public URL for TradingView:
ngrok http 5001
Copy the generated Forwarding URL (e.g. https://your-subdomain.ngrok-free.dev).
📊 TradingView Integration
Step 1: Save the Pine Script
Create a new Pine Editor script in TradingView and paste the code from ut_bot_strategy.pine. Click Save and Add to chart.
Step 2: Set Up the Alert
- Press
Alt + Ato open the Create Alert dialog. - Condition: Select
UT Bot Strategy – Buy & Sell with SL/TP. - Trigger: Select
Alert() function calls only(crucial for sending dynamic JSON parameters). - Webhook URL: Under the Notifications tab, check Webhook URL and paste your public ngrok URL with
/webhookat the end:https://your-subdomain.ngrok-free.dev/webhook - Alert Name:
UT Bot MT5 Automation. - Clear the Message box.
- Click Create.
🔒 Security Recommendations
- Demo First: Always test with a demo/trial account before using real money.
- Firewalling: Use an authentication mechanism or limit ngrok access if deploying to a production VPS.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。