Summary

This repository curates open-source projects leveraging Large Language Models (LLMs) for financial market research, trading strategy development, and execution. It targets developers, researchers, and power users seeking to integrate LLMs into automated trading systems. The primary benefit is a structured, categorized overview of the LLM-driven trading agent ecosystem, distinguishing itself from traditional quantitative finance libraries.

How It Works

The collection is organized into three core building blocks: Agents, Market Context Protocol (MCP) servers, and Skills. Agents are projects where LLMs actively participate in research or decision-making. MCPs act as middleware, enabling agents to interact with external tools, market data, or brokerages. Skills are reusable instructions and workflows that guide agents in performing specific trading tasks reliably. This modular design facilitates the construction of complex, LLM-powered financial trading systems.

Quick Start & Requirements

This repository is a curated list, not a single runnable project. Installation and requirements vary significantly per individual project listed. Some projects mention Docker for deployment (e.g., KylinMountain/TradingAgents-AShare). Users should consult the documentation of specific projects for prerequisites such as Python versions, GPU/CUDA requirements, API keys, or datasets. Links to official quick-start guides, documentation, and demos are often provided within individual project entries.

Highlighted Details

• Focuses exclusively on LLM-driven trading agents, MCPs, and skills, explicitly excluding traditional AI/ML finance tools.
• Selection criteria emphasize public code, clear LLM integration, recent activity, documentation quality, distinct roles, and visible adoption.
• Highlights key foundational projects like TauricResearch/TradingAgents (multi-agent debate framework), alpacahq/alpaca-mcp-server (brokerage integration), and tradermonty/claude-trading-skills (equity research workflows).
• Includes projects with multi-asset support (equities, crypto, forex, futures) and diverse LLM integrations (e.g., LangGraph, AutoGen, Claude Code, Gemini).

Maintenance & Community

The list is stewarded by the LLMQuant community. Contributions are welcomed via GitHub issues and pull requests, following guidelines in CONTRIBUTING.md. Links to the LLMQuant organization's Website, GitHub, and LinkedIn are provided for community engagement and updates.

Licensing & Compatibility

The awesome-trading-agents repository itself does not specify a license. Individual projects listed within the repository will have their own licenses, which potential adopters must independently verify for compatibility, especially concerning commercial use or integration into closed-source systems.

Limitations & Caveats

This list is narrowly scoped to LLM-native trading applications and does not cover traditional quantitative finance libraries, time-series models, or reinforcement learning trading bots. Users seeking those specific types of tools should refer to the related awesome lists mentioned in the README. The rapid evolution of LLM technology means project activity and relevance can change quickly.