This project provides an AI-powered framework for cryptocurrency trading, targeting quantitative traders and developers. It leverages a graph-based workflow architecture and LLM agents to generate data-driven trading signals and manage portfolios, aiming for sophisticated decision-making across multiple timeframes.

How It Works

The system utilizes LangGraph to construct a directed acyclic graph (DAG) of specialized nodes. This architecture processes market data through parallel, multi-timeframe analysis nodes, feeding into strategy nodes that apply technical indicators. Signals are aggregated with adaptive weighting, combined with risk parameters, and finally processed by LLM agents for portfolio management decisions. This compositional approach allows for flexible customization of data sources, timeframes, and strategies without core code modification.

Quick Start & Requirements

• Install: Clone the repository and use uv for dependency management (uv pip sync).
• Prerequisites: Python 3.9+ (3.12 recommended), Binance account with API keys. OpenAI or other LLM provider API keys are required for LLM integration.
• Setup: Copy config.example.yaml to config.yaml and populate .env with API keys.
• Docs: Project README

Highlighted Details

• AI-enhanced decision making via LLMs for portfolio management.
• Modular, node-based workflow architecture for flexibility.
• Ensemble approach aggregating multiple technical strategies.
• Multi-timeframe analysis for robust signal generation.
• Automatic generation of computational graph visualizations.
• Comprehensive backtesting capabilities with detailed metrics.

Maintenance & Community

No specific community channels (Discord/Slack) or notable contributors are mentioned in the README. The project appears to be actively developed by a single maintainer.

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

Live trading execution via a Binance gateway is planned but not yet fully implemented; the current focus is on signal generation. The README strongly advises thorough backtesting before deploying with real capital due to inherent financial risks.