Claude Skill

KylinMountain/TradingAgents-AShare

TradingAgents-AShare is an open-source multi-agent AI system for A-share investment research. 15 agents simulate institutional debate with full visualization, OpenClaw/Claude Code integration, and...

Overview

Stars638
Forks199
LanguagePython
Last pushed2026-06-30
Last synced2026-07-03
View on GitHub

Repository

OwnerKylinMountain
RepositoryTradingAgents-AShare
Full nameKylinMountain/TradingAgents-AShare
Repo ID1,170,370,148

Install this Skill

docker run -d -p 8000:8000 \

Registry

Typeopenclaw_skill
Quality score80/100
Verificationreadme_parsed
Last verified2026-06-09
Platforms
ClaudeOpenClaw
Capabilities
searchimageagentsaifinancellmopenclawopenclaw-skillsskillsstock
Detected files
README.mdpyproject.tomlrequirements.txttests
Config keys
TA_APP_SECRET_KEYDATABASE_URLTOKENYOUR_API_TOKEN
Install methods
  • docker run -d -p 8000:8000 \
  • git clone https://github.com/KylinMountain/TradingAgents-AShare.git
  • npm install

Summary

TradingAgents-AShare is an intelligent investment research system for A-shares, built on the TradingAgents architecture. It orchestrates 15 AI agents to simulate institutional collaboration and real-time debate, with full workflow visualization, OpenClaw / Claude Code integration, and one-click Docker deployment.

Chinese description

A股多智能体智能投研系统 — 基于TradingAgents架构,15名AI Agent模拟机构协作与实时辩论对抗,全流程可视化,支持OpenClaw / Claude Code集成,Docker一键部署。

Key features

  • 15 AI agents simulating institutional collaboration and real-time debate
  • Full workflow visualization for transparent decision-making
  • OpenClaw / Claude Code integration for extensible skill execution
  • One-click Docker deployment for rapid setup
  • Built on TradingAgents architecture tailored for A-share market

Use cases

  • Automated A-share investment research and analysis
  • Multi-agent debate for stock selection and risk assessment
  • Institutional-grade collaborative decision-making simulation
  • Visual workflow monitoring for research teams
  • Rapid prototyping of AI-driven trading strategies

README excerpt

# TradingAgents-AShare:A股智能投研多智能体系统 本项目是基于多智能体协作的 A 股深度分析系统,模拟顶级投研机构的决策闭环,通过 14 名专业 Agent 的多空辩论与风控博弈,为投资者提供结构化的交易建议。 [在线体验](https://app.510168.xyz) | [Releases](https://github.com/KylinMountain/TradingAgents-AShare/releases) | [OpenClaw 技能](https://clawhub.ai/kylinmountain/tradingagents-analysis) <div align="center"> <img src="assets/web/analysis.png" width="100%" alt="智能分析"/> <p><em>14 名智能体实时协作,左侧对话驱动,右侧可视化全流程</em></p> </div> > TradingAgents-AShare 已正式上线 OpenClaw!您只需通过 `tradingagents-analysis` 技能,即可让您的 AI助手具备专业的 A 股深度投研能力。 ## 功能特性 ### 辩论对战可视化 点击 Agent 卡片即可打开辩论 Drawer,实时观看多空对抗与风控三方辩论。垂直时间线按 Round 分组,Token 级流式呈现每位 Agent 的发言,裁决卡片独立高亮展示。 <div align="center"> <img src="assets/web/debate_drawer.png" width="80%" alt="辩论对战可视化"/> </div> ### 意图驱动的自然语言交互 直接输入"调研茅台短线"即可自动识别标的、解析投资周期,支持短线与中线双周期分析,无需填写表单。 ### 自选股与定时分析 数据库持久化自选列表,支持批量加入股票、自定义周期与触发时间,并可在前端批量更新、删除或手动测试定时任务。定时分析会自动复用持仓上下文,连续失败自动停用,无需人工干预。 <div align="center"> <img src="assets/web/timer_analysis.png" width="80%" alt="定时分析"/> </div> ### 持仓追踪与跟踪看板 支持导入持仓数据,自动记录持仓、成本价与仓位占比,并可一键将持仓标的补齐到定时分析列表。控制台会展示跟踪看板摘要,完整看板页支持查看实时价格、当日区间、持仓盈亏与上一交易日报告区间,方便盘中快速跟踪。 ### 结构化研报管理 分析结果结构化存储,支持按标的、日期检索历史研报,决策卡片一目了然地展示方向、置信度、目标价与止损价。 <div align="center"> <table style="width: 100%"> <tr> <td width="50%"><img src="assets/web/reports.png" alt="历史报告"/><br><em>研报历史</em></td> <td width="50%"><img src="assets/web/detail.png" alt="研报详情"/><br><em>深度详情</em></td> </tr> </table> </div> ### 多模型厂商支持 OpenAI、Anthropic、Google Gemini、DeepSeek、Moonshot、智谱、硅基流动等,用户可在前端自由切换模型厂商与具体模型;保存配置后会自动执行模型 warmup,也可以在设置页手动发送“你好”查看模型原始返回,便于排查接入问题。 <div align="center"> <img src="assets/web/settings.png" width="80%" alt="定时分析"/> </div> ## 核心架构 TradingAgents 模拟真实交易机构的部门协作,将复杂任务拆解为专业的智能体角色: <p align="center"> <img src="asset

Topics

Explore more

Data from GitHub. Synced on 2026-07-03