lzwme/finance-quant-skills

What it solves

This repository provides a collection of specialized "Skills" for AI agents (specifically Claude and Codex) to enable them to perform professional quantitative finance tasks. It bridges the gap between general-purpose LLMs and the specific technical requirements of quantitative trading, such as interacting with financial data APIs, using backtesting frameworks, and following specific trading platform documentation.

How it works

Following the Agent Skills standard, the project organizes specialized instructions, scripts, and resources into folders. Each "Skill" contains a SKILL.md file that tells the AI agent how to execute specific tasks. These skills can be installed into AI coding assistants like Claude Code or Codex via a marketplace system or CLI tools (npx skills), allowing the agent to dynamically load the necessary context and rules when a user asks for financial analysis or strategy development.

Who it’s for

Quantitative traders, financial analysts, and developers who use AI agents to automate the development of trading strategies, perform financial data analysis, or generate institutional-grade research reports.

Highlights

• Comprehensive Data Integration: Includes skills for multiple financial data interfaces like AKShare, BaoStock, Tushare, and jqdatasdk.
• Backtesting Support: Provides specialized skills for popular Python backtesting frameworks such as Backtrader and RQAlpha.
• ** uma Strategy Development**: Includes guidelines and API references for platforms like QMT and JoinQuant.
• Institutional Research: Features a skill for generating institutional-grade equity research reports.
• Multi-Agent Compatibility: Supports installation across different AI tools including Claude Code, Codex, and OpenClaw.