# dolphinquant/echolon

**Category:** 💰 Finance & Fintech  
**Repository:** https://github.com/dolphinquant/echolon  
**GitHub Stars:** 2  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/dolphinquant-echolon

## Description
LLM-agent-native backtest framework for SHFE daily futures research. 23 MCP tools (strategy validation, scaffolding, indicator catalog, error-code lookup), 22 in-package skills, 32 catalogued error codes, working strategy templates. pip install echolon then claude mcp add -s user echolon -- echolon-mcp.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `uvx` (confidence: high):

```json
"mcpServers": {
  "echolon": {
    "command": "uvx",
    "args": ["echolon"]
  }
}
```

## Documentation

## What dolphinquant/echolon MCP server does

The dolphinquant/echolon MCP server is an agent-oriented backtest framework focused on research with SHFE daily futures. Its MCP interface exposes 23 tools for common tasks in a strategy-development workflow. These include validating strategies, generating scaffolding, browsing an indicator catalog, and looking up documented error codes.

The package also contains 22 in-package skills. These skills provide packaged guidance or workflows for the framework, while the included strategy templates give agents starting points for creating research code. The project catalogs 32 error codes, which can help an agent interpret framework-specific failures during development or validation.

This is a specialized fit for SHFE futures research. The supplied information does not describe support for other exchanges, asset classes, live trading, broker connectivity, or production order execution.

## How it works

After installation, the framework exposes an `echolon-mcp` executable as an MCP server. An MCP-compatible client can register that executable and invoke its tools during an agent session. The provided setup example registers it for the user scope with Claude's MCP command:

```bash
pip install echolon
claude mcp add -s user echolon -- echolon-mcp
```

Once connected, an agent can use the server's research-oriented tools instead of treating the framework as an opaque code library. Strategy validation can be used to check work, scaffolding can help create a project structure, and the indicator and error-code catalogs can provide framework-specific reference information. The available material does not specify the transport details, individual tool names, input schemas, or output formats.

## Setup and configuration

Install the Python package named `echolon` with pip. The server command shown by the project is `echolon-mcp`. The example then adds a client entry named `echolon` at user scope through the `claude mcp add` command.

No environment variables, API keys, account credentials, or external service configuration are documented in the supplied material. The package name and server executable are the only required setup details stated here. Because the README content is unavailable, client-specific configuration beyond the shown Claude command cannot be confirmed.

## Tools and capabilities

The documented capabilities include:

- Strategy validation for checking futures research strategies.
- Scaffolding for creating the starting structure of a strategy project.
- An indicator catalog for discovering available indicators.
- Error-code lookup covering 32 cataloged codes.
- Twenty-two packaged skills for agent workflows.
- Working strategy templates that can serve as implementation references.

The listing reports 23 MCP tools in total, but does not enumerate every tool or identify which tools correspond to each packaged skill. It also does not state whether the framework fetches market data itself or expects data to be supplied by the user.

## Limitations and notes

The dolphinquant/echolon MCP server is described specifically for SHFE daily futures research, so its suitability for other markets or frequencies is not established. Nothing in the supplied description confirms live-market execution, broker integration, portfolio management, or automated deployment. The available information also does not identify a software license, supported MCP clients beyond the shown Claude registration command, or version compatibility requirements.

Use the server when the main requirement is agent-assisted strategy research and backtest development within its stated SHFE scope. Verify the repository documentation before relying on undocumented data sources, execution behavior, client support, or configuration options.

_Full upstream README: https://allmcps.com/mcp/dolphinquant-echolon/readme_