MCP Order Flow Server

Name: MCP Order Flow Server
Author: fintools-ai

@fintools-ai

代码仓库

关于 MCP Order Flow Server

This is the mcp to fetch and get order flow information

#mcp-order-flow #order-flow-analysis #options-trading

基本信息

分类

其他

运行时

python

传输方式

stdio

发布者

fintools-ai

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "mcp-order-flow-server": {
      "command": "python",
      "args": [
        "-m",
        "venv",
        "venv"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is MCP Order Flow Server?

MCP Order Flow Server provides real-time order flow analysis for options trading decisions. It consumes quote data from Redis and delivers structured analysis through MCP tools, including momentum indicators, size dynamics, pattern detection, support/resistance levels, and market behavior flags.

How to use MCP Order Flow Server?

Install dependencies, configure Redis environment variables (REDIS_HOST, REDIS_PORT, etc.), and run the server via python -m src.mcp_server or the dev_start.sh script. The primary tool is analyze_order_flow(ticker, history, include_patterns), which returns XML-formatted order flow data.

Key features of MCP Order Flow Server

Real-time momentum indicators (bid/ask lifts, drops, price changes)
Size dynamics analysis (volume patterns, large order detection)
Pattern detection (absorption, stacking, sweeps, momentum shifts)
Support/resistance level identification from order flow
Market behavior flags (YES/NO) for LLM interpretation
Background processing every 1 second for updated metrics

Use cases of MCP Order Flow Server

Enabling LLM agents to make informed options trading decisions
Detecting institutional accumulation or distribution via absorption patterns
Identifying trend initiation or reversal through momentum shifts
Recognizing aggressive execution or stop hunting with sweep detection
Providing structured order flow context for automated trading assistants

FAQ from MCP Order Flow Server

What are the runtime dependencies?

Python 3.10 or later, a running Redis server, and the mcp-trading-data-broker publishing quotes into Redis.

How is the data stored and for how long?

Raw quotes and detected patterns are stored in Redis with a default TTL of 3600 seconds. Calculated metrics have TTLs ranging from 60 to 600 seconds depending on the window.

What patterns does the server detect?

It detects absorption, stacking, momentum shifts, and sweeps. Each pattern has specific trigger conditions defined in the README (e.g., >15k size change for sweeps).

How do I configure the server?

Set environment variables such as REDIS_HOST, REDIS_PORT, REDIS_DB, LOG_LEVEL, and PROCESSOR_INTERVAL (default 1 second). A Claude Desktop configuration example is provided for integration.

Is authentication or authorization supported?

The README does not mention any authentication or authorization mechanisms for the MCP server or its Redis connection. An optional REDIS_PASSWORD environment variable exists for Redis authentication.