# AlgoRover Research

> AlgoRover Research is an MCP server for trading analysis. Search published trading strategies and run historical backtests via MCP. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Platforms:** API, CLI
- **Repository:** https://github.com/algorover/algorover-research
- **MCP Registry name:** io.github.algorover/algorover-research
- **Categories:** [trading analysis](https://ai-powered.com/categories/trading-analysis)
- **Listing updated:** 2026-11-19

## About

The README says AlgoRover Research is a hosted service reached through a remote MCP endpoint, so nothing needs to be installed locally beyond connecting the client. Its workflow covers finding published strategies, checking their versions and runnable presets, choosing a market and a 90-day or 365-day historical window, reviewing the configuration, then starting, monitoring or cancelling a backtest.

Results are tied to the signed-in account. Metrics, trade details, task history and research report downloads can be retrieved across conversations, and the backtests run on public market data inside the vendor's service rather than on the user's machine.

The server is aimed at traders and quantitative researchers who want to explore strategy performance conversationally. It does not connect exchange accounts or place orders, and the README stresses that historical results do not establish future returns.

## Features

- **Strategy catalog search:** Finds published trading strategies and shows their versions, package licences and runnable presets.
- **Backtest preparation:** Lets you pick a supported market and preset and review the configuration before starting a run.
- **Historical windows:** Runs backtests over a 90-day or 365-day historical period on public market data.
- **Task control:** Starts backtests, reports progress and cancels unfinished tasks.
- **Results and reports:** Returns metrics, trade details, task history and downloadable research reports for your account.
- **Data deletion on request:** Deletes a selected backtest result or all Research data, but only after separate authorization.

## Good for

- Comparing EMA pullback strategies before choosing one to test
- Running a 90-day backtest on an ETH preset and reviewing the configuration first
- Checking progress of a running backtest and cancelling it if needed
- Retrieving trade metrics and a report link from a past successful run

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http algorover-research https://research.algorover.ai/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://research.algorover.ai/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "algorover-research": {
      "url": "https://research.algorover.ai/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "algorover-research": {
      "type": "http",
      "url": "https://research.algorover.ai/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "algorover-research": {
      "serverUrl": "https://research.algorover.ai/mcp"
    }
  }
}
```

### Grok Bot (Grok Bot desktop app → Connect apps → Add a custom connector)

```sh
https://research.algorover.ai/mcp
```

Enter a Name (for example "AlgoRover Research"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "algorover-research": {
      "type": "http",
      "url": "https://research.algorover.ai/mcp"
    }
  }
}
```

## Tools

11 of 16 tools say they're read-only. 5 of 16 tools say they can delete or change data. From the server's own tool annotations; not verified by us.

- `get_profile` (says read-only): 读取当前已授权的 AlgoRover 账号资料。
- `search_strategies` (says read-only): 打开 AlgoRover Research 内嵌策略目录，搜索公开策略并返回精确版本。含连字符的完整策略 slug 精确匹配，普通关键词按相关性搜索。首次目录显式传入会话语言 locale：中文用 zh-CN，英文用 en；用户指定语言时优先采用。后续详情和目录保持相同语言。
- `get_strategy` (says read-only): 读取公司已发布策略及其现有解析状态、预设和市场。
- `prepare_strategy` (says it can change data): 为指定的已发布策略版本安排预设解析，不启动回测。
- `prepare_backtest` (says read-only): 准备固定版本、预设和历史窗口，返回预览；不创建回测。
- `get_strategy_download` (says read-only): 取得确定版本的公开策略下载信息；不安装或执行文件。
- `start_backtest` (says it can change data): 创建个人远端历史回测任务，返回任务标识。
- `get_backtest` (says read-only): 查询本人回测任务的状态、错误和保留期限。
- `list_backtests` (says read-only): 分页列出本人的历史回测任务。
- `cancel_backtest` (says it can change data): 请求取消本人回测。
- `get_backtest_result` (says read-only): 读取本人已完成回测的真实指标、有限曲线和复现信息。
- `list_backtest_trades` (says read-only): 分页读取本人已完成回测的成交明细。
- `get_backtest_report_download` (says read-only): 为本人已存在的报告文件生成短期下载地址。
- `delete_backtest_result` (says it can change data): 删除本人终态回测结果并撤销文件访问。
- `get_research_usage` (says read-only): 查询本人的运行中和排队任务数量。
- `delete_research_data` (says it can change data): 仅在用户明确要求时永久删除个人研究数据并撤销当前授权。删除期间暂停访问；完成后可使用同一账号重新登录授权，开始新的研究，旧数据不会恢复。

## FAQ

### Does it place live trades or connect to exchange accounts?

No. The README says AlgoRover Research supports historical research only and does not connect exchange accounts or execute trading orders.

### Which permissions does the default sign-in request?

The default command requests research:read and research:run. Deleting data needs the additional research:delete permission, which is not requested by default.

### Which historical windows can be used?

Backtests can use a 90-day or 365-day historical window.

### How is the server connected?

The registry entry lists a streamable-http remote at https://research.algorover.ai/mcp, and the README describes installing it as a Codex plugin with OAuth sign-in.

### How do I add AlgoRover Research to Claude Code?

Run this command in your terminal: claude mcp add --transport http algorover-research https://research.algorover.ai/mcp

### What are the best alternatives to AlgoRover Research?

The closest alternatives to AlgoRover Research are Universal Quant Tournament MCP, BaseEdge Labs Account, Tactfolio, PatternsWizard pattern statistics and Onv MCP.

## Where this comes from

- Description: written from the project's README and its MCP Registry entry.
- Status and tools: checked automatically by AIPoweredBot on 20 November 2026.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [Universal Quant Tournament MCP](https://ai-powered.com/mcp/universal-quant-tournament-mcp): Universal Quant Tournament MCP is an open-source MCP server for trading analysis.
- [BaseEdge Labs Account](https://ai-powered.com/mcp/baseedge-labs-account): BaseEdge Labs Account is a freemium MCP server for trading analysis.
- [Tactfolio](https://ai-powered.com/mcp/tactfolio): Tactfolio is a freemium MCP server for trading analysis.
- [PatternsWizard pattern statistics](https://ai-powered.com/mcp/patternswizard-pattern-statistics): PatternsWizard pattern statistics is a free MCP server for analytics and data.
- [Onv MCP](https://ai-powered.com/mcp/onv-mcp): Onv MCP is a freemium MCP server for crypto and web3.
- [AStock Quant Toolkit](https://ai-powered.com/mcp/astock-quant-toolkit): AStock Quant Toolkit is an open-source MCP server for trading analysis.
- [Backtester](https://ai-powered.com/mcp/backtester): Backtester is an MCP server for trading analysis.
- [Strategy Lab by Stocks Developer](https://ai-powered.com/mcp/strategy-lab-by-stocks-developer): Strategy Lab by Stocks Developer is a free MCP server for finance and payments.

---
Source: https://ai-powered.com/mcp/algorover-research
