# Algovesta Trading MCP server

Source: https://glimind.com/server/mcp-registry%2Fcom.algovesta%2Ftrading
Registry id: `mcp-registry/com.algovesta/trading`
Status: auth-required (checked 2026-10-26T23:05:51.524Z)

## Measurements (30 days)

- Uptime: 100%
- Checks: 213
- Outages (90 days): 0

## Install

```json
{
  "mcpServers": {
    "algovesta-trading-mcp-server": {
      "type": "http",
      "url": "https://api.algovesta.com/mcp"
    }
  }
}
```

Claude Code:

```sh
claude mcp add --transport http algovesta-trading-mcp-server https://api.algovesta.com/mcp
```

## Tools (20)

- `backtest_my_signals`: Replays your own past signals with different settings (queued job)
- `cancel_order`: Cancels a pending order
- `close_position`: Closes fully or partially (MT5: full close only)
- `compare_venues`: Ranks your connected exchanges on measured price and spread
- `compile_policy`: Turns plain-language rules into a policy preview
- `create_strategy`: New strategy, always created with real money OFF
- `get_job_status`: Progress and result of a queued job
- `get_market_price`: Live price with freshness reported
- `get_portfolio_context`: Every connected account in one call
- `get_trade_history`: Closed trades and performance across crypto, MT5 and paper
- `import_tradingview_backtest`: Recomputes a TradingView trade export with real fees and slippage (queued job)
- `list_open_orders`: Pending limit orders (paper book)
- `list_strategies`: Your TradingView strategies and whether real money is on
- `modify_position`: Moves stop-loss and take-profit
- `place_order`: Opens a position (live crypto: market orders only)
- `replay_channel`: Backtests a Telegram channel against your rules
- `simulate_order`: Dry-run including the policy verdict
- `simulate_policy`: Applies a risk policy to the trades you actually closed (queued job)
- `update_strategy`: Changes strategy settings (never auto_trade)
- `verify_receipt`: Checks signature and hash chain

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.