LLMQuant/awesome-trading-agents

What it is:

Awesome Trading Agents is a curated list (an “awesome list”) of open-source projects where large language models (LLMs) help with market research, trading decisions, or connecting AI agents to market data and execution tools. It’s maintained by the LLMQuant community.

What it collects:

The list focuses on three building blocks:

• Agents – projects where an LLM is part of the actual research or trading decision (e.g., analyst teams, single-agent traders, research copilots, live trading experiments, benchmarks, and tools that write or improve strategies).
• MCPs – servers that let an agent call external tools through the Model Context Protocol (a standard way for AI agents to access data or actions). These cover market data, broker/exchange actions, research tools, and backtesting services.
• Skills – (implied from the table of contents) collections of capabilities or instructions for agents, such as equity research, crypto/DeFi, strategy coding, and brokerage execution.

How it’s organized:

The README is divided into sections: Agents (with subcategories like multi-agent systems, single-agent traders, research copilots, real-money experiments, prediction-market specialists, benchmarks, and strategy-coding agents), MCPs (market data, brokerage, research tools, TradingView bridge, prediction markets, backtesting platforms), Skills (equity research, crypto/DeFi, strategy coding, brokerage), and Resources (papers and learning material). Each entry is a link to a GitHub project with a short description.

Selection criteria:

Projects are chosen if they have public code or artifacts, clear LLM-driven behavior, recent activity, useful documentation, a distinct role, and visible adoption. The list explicitly excludes classic quant libraries, time-series models, and reinforcement-learning trading bots (those are covered by other lists).

Notable highlights:

The README suggests three standout projects per category:

• Agents: TauricResearch/TradingAgents (multi-agent debate framework), virattt/ai-hedge-fund (LLM-driven equity trading with analyst personas), HKUDS/AI-Trader (agent-native trading platform).
• MCPs: alpacahq/alpaca-mcp-server (brokerage), krakenfx/kraken-cli (exchange), financial-datasets/mcp-server (market data).
• Skills: tradermonty/claude-trading-skills, himself65/finance-skills, RKiding/Awesome-finance-skills.

Other details:

The list is bilingual (English and Simplified Chinese), licensed under CC0-1.0, and includes a note that dates are not shown for every item but activity is still checked before adding or updating entries. It also points to related awesome lists for broader AI-in-finance and quant resources.

In short, it’s a well-organized, community-curated directory for anyone interested in using LLMs for trading and financial analysis, with a focus on practical, actively maintained open-source projects.