ArvinLovegood/go-stock

🦄🦄🦄AI赋能股票分析：AI加持的股票分析/选股工具。股票行情获取，AI热点资讯分析，AI资金/财务分析，涨跌报警推送。支持A股，港股，美股。支持市场整体/个股情绪分析，AI辅助选股等。数据全部保留在本地。支持DeepSeek，OpenAI， Ollama，LMStudio，AnythingLLM，硅基流动，火山方舟，阿里云百炼等平台或模型。

What is go‑stock?

go‑stock is a desktop application (built with the Go‑based Wails framework and NaiveUI) that lets users look at real‑time market data for Chinese A‑shares, Hong‑Kong stocks, U.S. stocks and ETFs, and then ask a large‑language‑model (LLM) to analyse the data. The tool bundles a huge amount of finance‑related “AI data tools” (150+), a skill/plug‑in system, and several AI‑agent modes (React, PlanExecute, DeepAgents) that can run prompts, call external tools, and even understand images of stock charts.

Core capabilities

Area What the app does
Market data Live quotes for A‑share, HK, US markets, ETF search, fund rankings, K‑line (candlestick) charts with 13 drawing tools, volume distribution, wave theory, auto‑divergence detection, etc.
AI‑powered analysis Users can type or continue a conversation with an LLM (OpenAI, Ollama, LMStudio, DeepSeek, GLM, etc.). The AI can generate stock reports, suggest picks, answer follow‑up questions, and even analyse uploaded chart screenshots.
Agent / skill system Three built‑in agent modes (React, PlanExecute, DeepAgents). Agents can call over 150 built‑in data‑tool functions, use a “skill” plug‑in architecture, and keep a long‑term memory vector store (chromem‑go).
Back‑testing & strategy Prompt‑template back‑testing, recommendation back‑testing (multiple holding periods), daily‑review and pre‑market strategy generation, performance metrics (win‑rate, Sharpe, draw‑down, etc.).
Alerts & automation Real‑time signal monitoring, audio/voice alerts, scheduled tasks, push notifications to Feishu or DingTalk, VIP‑only background monitoring.
Extensibility MCP tool server integration, OAuth, skill marketplace, API for prompt‑template lookup, dynamic tool search.

How it works

  1. UI – A native desktop UI rendered with Vue + NaiveUI, packaged via Wails for Windows (primary target) and macOS (universal binary).
  2. Backend – Go services fetch market data (Tushare, proprietary APIs), store user settings, run scheduled jobs, and host the LLM‑agent runtime.
  3. LLM integration – The app can call any OpenAI‑compatible endpoint or local model server (Ollama, LMStudio, AnythingLLM, etc.). Visual‑understanding is supported by routing images to models that accept vision inputs.
  4. Tool calling – Agents use the Eino ADK framework to invoke built‑in “tools” (e.g., GetBkFundFlowRank, GetClsCalendar) and custom skill packages. Results are fed back into the conversation.
  5. Persistence – SQLite stores signal logs, back‑test results, and a chromem‑go vector store for knowledge‑base and long‑term memory.

Who might use it?

  • Individual traders who want a single desktop client that combines market data, charting, and AI‑driven commentary.
  • Hobbyist quant developers looking for a plug‑in friendly environment to experiment with LLM‑based stock‑selection strategies.
  • Anyone interested in exploring AI agents that can browse financial data, run back‑tests, and generate daily reports.

Limitations

  • The README stresses that the software is for entertainment only; AI stock advice is not guaranteed and should not be used for real investment decisions.
  • Primary development is on Windows 10; macOS support exists but may be limited.
  • Many advanced features (background monitoring, skill marketplace, VIP‑only tools) require a paid sponsorship tier.

TL;DR: go‑stock is a Windows/macOS desktop app that merges live stock market data with large‑language‑model agents, offering chart analysis, AI‑generated reports, back‑testing, and extensible skill plug‑ins—all built with Go, Wails, and Vue/NaiveUI.

Key links from the README

  • Quick‑start & full manual: docs/快速开始指南.md, docs/go-stock使用手册.md
  • Download: go-stock-windows-amd64.exe, go-stock-darwin-universal
  • Supported LLM platforms: OpenAI, Ollama, LMStudio, DeepSeek, GLM, ByteDance, Alibaba, Moonshot, Tencent, iFlytek, MiniMax, Xiaomi, SiliconFlow, etc.

Note: The project is actively maintained (changelog up to Sep 2026) and includes a roadmap of future AI‑agent and visual‑understanding features.

License & community

  • Open‑source on GitHub (stars badge shown). The repo includes a QQ group for users and a sponsorship plan for additional features.

Bottom line: This is a genuine software project that sits squarely in the AI‑augmented finance tooling space.

Generated by AI‑curator

Metadata

  • Stars: shown in badge
  • Latest release: v2025.2.12.7‑alpha (as of README) but changelog shows later 2026 releases.
  • Primary language: Go (backend) + Vue/JS (frontend)
  • License: not explicitly listed in README.

End of explanation

Related

  • Project
    hengruiyun/AI-Stock-Master

    AI Stock Master is a Windows/macOS desktop app that blends quantitative stock‑analysis algorithms (RTSI, TMA, MSCI) with a locally‑run LLM (Mini Ollama) to generate natural‑language market reports and trading signals for Chinese, Hong‑Kong and US equities. It’s a research‑oriented tool, not a commercial trading platform.

  • Project
    EthanAlgoX/AIStock

    AI Stock is an agentic research workspace that helps investors analyze stocks, compare investment perspectives, and simulate trading strategies using LLM-powered tools.

  • Project
    jundizhou/easy-stock

    An AI-powered research workstation for A-share investors that transforms market data and influencer viewpoints into structured, evidence-based investment analysis.

  • Project
    YUHAI0/fin-agent

    An AI-powered financial analysis assistant that uses LLMs and Tushare data to enable natural language stock screening, strategy backtesting, and portfolio monitoring.

  • Project
    YoungCan-Wang/WyckoffTradingAgent

    WyckoffAgent is an open‑source, AI‑driven trading assistant that lets you converse with a “Wyckoff master” to analyse Chinese, Hong‑Kong and US stocks. It pulls real‑time data from TickFlow, identifies Wyckoff structures, generates LLM‑based research reports, and presents results via a React web UI, a terminal dashboard, or an Electron desktop app. All user data (portfolio, signals, back‑tests) lives in Supabase, and write actions go through an approval queue. The project is AGPL‑3.0, installable via Homebrew or pip, and includes detailed cost and architecture docs.