Vibe-Trading

Active
GitHub Python MIT

Description

Vibe-Trading is an open-source personal trading agent stack that combines LLMs, MCP tooling, and multi-agent workflows for algorithmic trading and backtesting.

Key Features

  • One command empowers your trading agent with full quantitative trading capabilities via LLM + MCP toolchain
  • Backtesting engine supporting 7 data sources (Tushare, OKX, CCXT, AkShare, etc.) with local caching
  • Connector-first architecture supporting IBKR, Robinhood, Tiger, Alpaca, OKX, Binance paper trading
  • Multi-agent collaborative workflows with Swarm mode, research goal runtime, and retry mechanisms
  • Three interaction modes: CLI, REST API, and MCP with 36+ MCP tools available
  • Built-in safety: mandate management, order guards, audit ledger, and instant circuit breaker

Use Cases

πŸ’‘ Personal quant traders quickly build AI-powered strategy backtesting and live trading systems
πŸ’‘ Describe trading strategies in natural language to auto-generate and backtest quantitative signal engines
πŸ’‘ Safely test multi-broker strategies using paper accounts with zero real capital risk
πŸ’‘ Build multi-agent collaborative research workflows for automated market data collection and analysis
πŸ’‘ Integrate via MCP protocol into other AI toolchains to extend trading capabilities

Tags

ai-agent algorithmic-trading backtesting fintech llm mcp mcp-tools multi-agent

Categories

πŸ”Œ MCP Tools 🧠 Multi-Agent

Quick Start

Install with pip install vibe-trading-ai. Use vibe-trading connector list to see available brokers, vibe-trading connector use <name> to select one. Run backtests with vibe-trading backtest or start the interactive CLI with vibe-trading cli. Docs at vibetrading.wiki/docs.
Visit GitHub Visit Website

Related Projects