simonlin1212/a-stock-data

What it solves

It provides a comprehensive, zero-authentication data toolkit for A-share (Chinese stock market) data. It eliminates the need for developers and AI agents to manually handle complex API parameters, authentication headers, and rate-limiting logic across dozens of fragmented financial data sources, consolidating them into a single, AI-ready interface.

How it works

The project is structured as a "Skill" file (structured Markdown with embedded Python) designed for direct injection into AI programming assistants like Claude Code, Codex, or OpenClaw. It implements an 11-layer architecture that fetches data from 19 different sources (including mootdx, Tencent, Eastmoney, and official exchange APIs) using a mix of TCP and HTTP requests. To prevent IP bans, it prioritizes non-blocking sources (like mootdx and Tencent) and uses a built-in rate-limiting wrapper (em_get()) for more restrictive sources like Eastmoney.

Who it’s for

• AI Developers: Those building financial agents or investment research tools who need reliable A-share data without managing multiple API keys.
• AI Programming Assistants: Users of Claude Code or similar tools who want to enable their agent to perform real-time stock valuation, thematic research, and macro analysis.
• Quantitative Researchers: Individuals needing a streamlined way to pull K-lines, reports, and capital flow data for the Chinese market.

Highlights

• Massive Coverage: 54 endpoints across 11 layers, covering everything from real-time quotes and K-lines to research reports, capital flows, and macro indicators (PMI, Social Financing).
• Zero Auth: Most data sources are completely free and require no registration or API keys (except for iwencai semantic search).
• Resilient Architecture: Includes a "backup source" strategy to automatically downgrade to alternative data providers if primary sources are blocked.
• AI-Native Design: Specifically formatted as a Skill file for seamless integration with LLM-based coding agents.
• Broad Data Scope: Includes specialized data like ETF options (Greeks/IV), shareholder counts, and historical valuation (PE/PB/PS) dating back to 2016.