jundizhou/easy-stock

What it solves

It addresses the difficulty of understanding the underlying structure of the A-share market (such as main themes, sentiment cycles, and trend validity) which traditional trading software—focused primarily on price and volume—often fails to capture. It transforms fragmented, unstructured data from various sources into a cohesive, evidence-based research system for individual investors.

How it works

The application uses a "local-first" architecture combining a Go backend, React/TypeScript frontend, and Electron desktop shell. It employs a specialized AI agent platform called Hermes to route tasks to various LLMs (like DeepSeek, OpenAI, or Claude). The system aggregates market data and news from sources like East Money and Caixin, and uses an agent-browser to automatically scrape and summarize viewpoints from market influencers on platforms like Xueqiu and Taoguba. This data is then processed through domain-specific models to analyze sentiment, trend ladders, and portfolio health.

Who it’s for

Individual investors in the A-share market who want to move beyond simple price tracking to a more structured, AI-assisted research and review process.

Highlights

• Automated Review: Automatically collects and synthesizes consensus and divergence from multiple market influencers.
• Sentiment & Trend Analysis: Analyzes "limit-up" ladders, sentiment cycles, and theme radars to identify market mainlines.
• AI Stock Analysis: Evaluates individual stocks based on paths like emotional limit-up or trend growth, providing customized reports.
• Portfolio Inspection: Performs health checks on holdings to identify concentration risks, style mismatches, and linkage risks.
• Knowledge Base: Allows users to build a long-term research "flywheel" by saving articles, analysis records, and trading mastery materials locally.