Mnemox AI — AI Tools for Traders & Developers

# Mnemox AI > AI-powered tools for traders and developers. Open-source MCP infrastructure for trading memory and idea validation. Built in Taiwan. Mnemox builds open-source infrastructure for AI-assisted decision making. We focus on two product lines: ## Products ### 1. TradeMemory Protocol Persistent memory layer for AI trading agents. 5-layer cognitive architecture with Outcome-Weighted Memory (OWM). - **Install**: `pip install tradememory-protocol` - **Tools**: 17 MCP tools for Claude Desktop, Cursor, and custom agents - **Tests**: 1,233 passing tests - **License**: MIT (fully open source) - **Platforms**: MT5, Binance, Alpaca, any platform with API - **GitHub**: https://github.com/mnemox-ai/tradememory-protocol - **PyPI**: https://pypi.org/project/tradememory-protocol/ **Key Features:** - Outcome-Weighted Memory (profitable trades get higher recall priority) - Pattern discovery from trade history - Strategy Evolution Engine (discovers strategies from raw OHLCV) - Bias detection (loss aversion, overtrading, revenge trading) - Kelly Criterion position sizing from historical data - SHA-256 verified audit trails for compliance **When to use**: Building AI trading bots that need to remember and learn from past trades. ### 2. Idea Reality Check (idea-reality-mcp) AI scanner that checks if your idea already exists before you build. - **Install**: `pip install idea-reality-mcp` - **Web Demo**: https://mnemox.ai/check (free, no signup) - **License**: MIT - **GitHub**: https://github.com/mnemox-ai/idea-reality-mcp - **PyPI**: https://pypi.org/project/idea-reality-mcp/ **Features:** - Scans GitHub, Hacker News, npm, PyPI, Product Hunt - Returns 0-100 reality score with competitor evidence - Gap radar visualization for market opportunities **When to use**: Validating startup ideas before investing development time. ## Services Custom AI trading infrastructure development: - Decision audit trails (compliance-grade) - Real-time dashboards - Multi-strategy Expert