YoungCan-Wang/WyckoffTradingAgent

What it solves

WyckoffAgent is an AI-driven trading assistant designed to automate the complex process of Wyckoff volume-price analysis. It bridges the gap between natural language queries and technical market data, allowing traders to perform market screening, portfolio diagnostics, and research report generation without manually analyzing every chart.

How it works

The system integrates real-time market data (via TickFlow) with Large Language Models (LLMs) to create an automated pipeline. It uses a "funnel" screening process to identify conceptual mainlines and buy points across A-shares, Hong Kong stocks, and US stocks. The agent can be triggered via a CLI, Web App, or Desktop application, and it supports the Model Context Protocol (MCP) to integrate with external tools like GitHub or local file systems.

Who it’s for

It is designed for traders and investors who follow the Wyckoff method of volume-price analysis and want to leverage AI to automate their research, screening, and risk management workflows.

Highlights

• Conversational Agent: Use natural language to trigger diagnostics, screening, and research reports.
• Multi-Market Support: Independent workflows for A-shares, Hong Kong stocks, and US stocks.
• AI-Powered Research: Generates reports categorized by logic (e.g., "Logic Bankruptcy" or "Jump-off Board").
• Automated Funnel Screening: Dynamically discovers mainlines and confirms buy points across thousands of stocks.
• Agentic Memory: Implements a layered memory system (L1 Atomic, L2 Scenario, L3 Persona) for better context recall.
• MCP Integration: Exposes 18 tools via the Model Context Protocol for use in tools like Claude Code or Cursor.
• Comprehensive Tooling: Includes a local dashboard, backtesting grids, and multi-channel notifications (Feishu, DingTalk, Telegram).