Tradememory Protocol
@mnemox-ai
Structured 3-layer memory system for AI trading agents. L1 records every trade automatically. L2 discovers patterns via reflection engine. L3 evolves strategy in real-time. Built on MCP. Supports MT5, Binance, Alpaca. Python/FastAPI/SQLite. MIT licensed.
Overview
What is Tradememory Protocol?
Tradememory Protocol is a structured 3-layer memory system for AI trading agents. It automatically records every trade (L1), discovers patterns via a reflection engine (L2), and evolves strategy in real-time (L3). Built on MCP, it supports MT5, Binance, and Alpaca exchanges. Implemented in Python/FastAPI with SQLite storage. MIT licensed.
How to use Tradememory Protocol?
As an MCP server, Tradememory Protocol is invoked by compatible MCP clients. The implementation uses Python, FastAPI, and SQLite. It supports configuration for MetaTrader 5 (MT5), Binance, and Alpaca exchanges. No specific commands or config keys are documented in the README.
Key features of Tradememory Protocol
- 3-layer memory for trading agents
- L1 records every trade automatically
- L2 discovers patterns via reflection engine
- L3 evolves strategy in real-time
- Supports MT5, Binance, Alpaca exchanges
- Built on MCP with Python/FastAPI/SQLite
Use cases of Tradememory Protocol
- Automatically record every trade made by an AI trading agent
