Backtest360 MCP server
{
"mcpServers": {
"backtest360": {
"command": "uvx",
"args": [
"backtest360-mcp"
],
"env": {
"BACKTEST360_API_KEY": "<backtest360_api_key>",
"BACKTEST360_ENGINE_URL": "https://..."
}
}
}
}{
"mcpServers": {
"backtest360": {
"command": "uvx",
"args": [
"backtest360-mcp"
],
"env": {
"BACKTEST360_API_KEY": "<backtest360_api_key>",
"BACKTEST360_ENGINE_URL": "https://..."
}
}
}
}{
"mcpServers": {
"backtest360": {
"command": "uvx",
"args": [
"backtest360-mcp"
],
"env": {
"BACKTEST360_API_KEY": "<backtest360_api_key>",
"BACKTEST360_ENGINE_URL": "https://..."
}
}
}
}{
"servers": {
"backtest360": {
"command": "uvx",
"args": [
"backtest360-mcp"
],
"env": {
"BACKTEST360_API_KEY": "${input:backtest360_api_key}",
"BACKTEST360_ENGINE_URL": "${input:backtest360_engine_url}"
}
}
},
"inputs": [
{
"id": "backtest360_api_key",
"type": "promptString",
"description": "Backtest360 api key",
"password": true
},
{
"id": "backtest360_engine_url",
"type": "promptString",
"description": "Backtest360 engine url",
"password": true
}
]
}{
"mcpServers": {
"backtest360": {
"command": "uvx",
"args": [
"backtest360-mcp"
],
"env": {
"BACKTEST360_API_KEY": "<backtest360_api_key>",
"BACKTEST360_ENGINE_URL": "https://..."
}
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsoninstall uvx backtest360-mcpversion 0.5.0
What it does
This server exposes the Backtest360 backtesting engine API as tools for AI agents. Users can discover indicators, build and validate trading strategies, run historical backtests, and analyze results through a conversational interface.
- Discover available indicators and their parameter schemas, then construct a trading strategy by combining them.
- Validate a strategy document for structural and logical errors before running a backtest against historical price data.
- Run a backtest on a strategy and compare its performance against buy-and-hold or other strategies on the same data.
Tools this server exposes
| Tool | Does |
|---|---|
| engine_info | Retrieve engine version, API contract, and health status. |
| get_me | Check the capabilities and limits of the configured API key, including permission scopes and current usage. |
| get_catalog | Fetch reference catalogs such as operators, execution modes, stop types, sizing methods, and bar frequencies. |
| list_indicators | List available indicators and their parameter schemas for strategy construction. |
| list_templates | Discover predesigned strategy templates and fetch one in full. |
| get_strategy_schema | Retrieve the JSON Schema for strategy documents. |
| validate_strategy | Validate a strategy without running it, returning structured errors with locations. |
| run_backtest | Execute a historical backtest on a strategy, with configurable detail levels and series downsampling. |
| get_latest_signal | Evaluate a strategy on the most recent bar only without computing P&L. |
| compare_backtests | Run multiple strategies on the same data for side-by-side comparison. |
| compute_stats | Calculate metrics from an externally produced returns series. |
| search_tickers | Search for asset symbols available for server-side data fetching. |
| list_tickers | List available asset symbols for data fetching. |
| get_data_range | Check available historical data and estimate bar count for a symbol. |
| get_ticker_info | Retrieve symbol identity and data coverage information. |
| get_quote | Fetch the latest available price for a symbol (paid plan only). |
| get_price_history | Retrieve OHLCV price history over a date range (paid plan only; long histories are downsampled). |
| list_macro_series | List available macroeconomic data series. |
| get_macro_series | Fetch observations from a macroeconomic data series. |
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.