Cpzai MCP server
{
"mcpServers": {
"cpzai": {
"type": "http",
"url": "https://mcp.cpz-lab.com/mcp"
}
}
}{
"mcpServers": {
"cpzai": {
"type": "http",
"url": "https://mcp.cpz-lab.com/mcp"
}
}
}{
"mcpServers": {
"cpzai": {
"type": "http",
"url": "https://mcp.cpz-lab.com/mcp"
}
}
}{
"servers": {
"cpzai": {
"type": "http",
"url": "https://mcp.cpz-lab.com/mcp"
}
}
}{
"mcpServers": {
"cpzai": {
"type": "http",
"url": "https://mcp.cpz-lab.com/mcp"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 1.1.0
What it does
MCP server that connects AI agents to CPZAI for quantitative trading. Exposes tools for strategy management, backtesting, order placement, portfolio tracking, and risk analytics across multiple brokers (Alpaca, Interactive Brokers, FIX). Intended for developers building trading automation with Claude, Cursor, GPT, or other MCP-compatible agents.
- Build and backtest trading strategies through an AI agent by creating strategies and retrieving backtest results.
- Execute live or paper trading orders and monitor positions across connected brokerage accounts.
- Analyze portfolio risk metrics like Value at Risk, Sharpe ratio, and drawdown on demand or through historical snapshots.
Tools this server exposes
| Tool | Does |
|---|---|
| list_strategies | List trading strategies with filtering options. |
| get_strategy | Retrieve details of a specific trading strategy by ID. |
| create_strategy | Create a new trading strategy. |
| update_strategy | Modify an existing trading strategy. |
| get_backtest_results | Retrieve backtest run results for strategies. |
| list_orders | List trading orders with filtering options. |
| place_order | Place a new trading order. |
| list_positions | List current portfolio positions across accounts. |
| sync_portfolio | Trigger portfolio synchronization across connected brokers. |
| list_accounts | List connected trading accounts. |
| get_market_data | Fetch real-time market quotes including price, bid/ask, and volume. |
| compute_risk | Calculate current risk metrics including Value at Risk, Sharpe ratio, and drawdown. |
| list_risk_snapshots | Retrieve historical risk snapshots. |
| execute_strategy | Execute a strategy on the Python backend. |
| list_webhooks | List webhook subscriptions. |
| create_webhook | Subscribe to platform events via webhook. |
| delete_webhook | Remove a webhook subscription. |
| get_profile | Retrieve the authenticated user profile. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.