trading
12 repositories
Simulated trading floor of specialized LLM agents — TradingAgents assigns LLM agents distinct roles—fundamental, sentiment, technical analysts, researchers, traders, risk managers—mirroring a real trading firm. Backed by an arXiv paper, it now runs on a personal ChatGPT Plus subscription as of v0.7.0, lowering the barrier for researchers and quant hobbyists to experiment with multi-agent trading strategies.
A One-Command MCP Toolkit That Turns Any Agent Into a Trader — Vibe-Trading equips AI agents with backtesting, quantitative analysis and algorithmic trading tools via MCP, installed with a single command. Built on Python, FastAPI and React 19, it targets developers who want to add real trading capabilities to their own agent setups rather than build everything from scratch.
A Rust core with Python bindings for identical backtest-to-live trading — NautilusTrader pairs a Rust engine with Python APIs so strategies run identically in backtesting and live trading, removing the usual gap between research and production code. It targets crypto, equities, FX, futures, options and sports betting, and suits quants and developers who need deterministic, event-driven execution at scale.
An AI-connected harness turns Hummingbot into an agentic trading platform — Hummingbot is an open-source framework for automated crypto trading bots across many exchanges, with users reporting over $34B in volume. Its new Condor harness links LLM-driven decision-making to deterministic trade execution via API, controllable through Telegram or a web dashboard, while the hbot CLI handles non-interactive bot control and monitoring.
Python quant trading framework covering stocks, futures, options and bitcoin — A Chinese open-source quantitative trading and investment framework supporting stocks, options, futures and bitcoin, with machine learning integration. It ships with tutorials, a non-coding UI, companion book code, and a live site offering AI research reports on major indices and assets.
A rebuilt quant stock screener for China's A-share market — Sequoia-X V2 is a from-scratch rewrite of an A-share screening system, using OOP design, vectorized computation and incremental data updates. It pulls adjusted daily K-line data via baostock into local SQLite (avoiding anti-scraping issues), runs six built-in strategies like TurtleTrade and RPS breakout, and pushes results to Feishu after market close.
Autonomous AI agents run your hedge fund on Solana — AutoHedge chains specialized AI agents—Director, Quant, Risk, and Execution—into a pipeline that generates trade theses, sizes risk, and executes orders autonomously. It currently trades live on Solana with Coinbase support coming soon, aimed at developers building automated quant trading systems.
An unofficial TypeScript rewrite to pull live TradingView data — V4 is a breaking rewrite of this community library in TypeScript, replacing the old Client API with a simpler getCandles/data API while still covering every v3 feature, per the included migration guide and coverage matrix. Useful for developers building charts, indicators, strategies or watchers on stock and crypto data without an official TradingView API.
A Claude-powered trading agent spanning prediction markets and crypto — CloddsBot is a self-hosted AI agent that trades across Polymarket, Kalshi, Binance, Hyperliquid, Solana DEXs and 5 EVM chains with 121+ skills. Built on Claude for autonomous edge-scanning and risk management, it also supports machine-to-machine payments via an agent commerce protocol, aimed at developers wanting a single terminal for cross-market trading.
An open-source, pluggable WebGL2 chart engine for trading apps — Vela is the open-source core behind LuxAlgo's chart product: a headless financial charting engine with a native WebGL2 renderer (canvas2d fallback) and swappable data providers, scripting engines and renderers via a plugin SDK. Useful for developers building custom trading or indicator-driven web apps who want full code control plus an optional ready-made workspace UI.
TradingView MCP server: search indicators and strategies, fetch Pine Script source. Works with Claude Code, Claude Desktop and Cursor. No API key needed.
A common harness for TradingAgents, AI Hedge Fund and your own trading bots — Glassbench plugs multi-agent finance systems like TradingAgents, AI Hedge Fund and Vibe-Trading into one adapter, running each on the same stock and date and logging every agent run to a shared database. It adds a live front end, LLM/framework benchmarks, backtesting, and Interactive Brokers paper trading, useful for anyone comparing agent trading strategies head to head.