Skip to content

StockAI v0.1.0

Choose a tag to compare

View all tags
@github-actions github-actions released this 06 Apr 14:41
v0.1.0
2a5eaa8

🎉 StockAI v0.1.0 — First Public Release

AI-powered stock sentiment analysis, right on your desktop.
Paste a ticker. Get deep, LLM-driven insights in seconds — no cloud subscription required.

StockAI Dashboard

✨ What is StockAI?

StockAI is a native desktop app that combines real-time news scraping with large language model analysis to give you actionable sentiment signals for any stock. It runs fully on your machine — bring your own OpenAI-compatible API key, or use a local Ollama model for complete privacy.

Think of it as a Bloomberg terminal assistant that works with both GPT-4o and your locally-running Qwen/Llama/Mistral model.

🚀 Highlights in v0.1.0

🔍 Real-Time News Intelligence

  • Automatically scrapes the latest news for any stock ticker from multiple sources (Google News, Yahoo Finance)
  • Extracts full article content — not just headlines — for deeper context
  • Powered by Playwright headless browser for reliable, JavaScript-rendered page scraping

🤖 Dual AI Provider Support

Provider Use Case
OpenAI / any OpenAI-compatible API Cloud-based analysis with GPT-4o, Claude-via-proxy, DeepSeek, etc.
Ollama (local) 100% private, offline analysis with Qwen 3.5 27B, Llama 3, Mistral, etc.

Switching providers auto-fills the correct endpoint URL — zero manual configuration.

📊 Interactive Dashboard

  • Clean, modern dark UI built with React + Tailwind CSS
  • Step-by-step progress visualization: Scraping → Extracting → Analyzing → Done
  • Structured analysis output with sentiment score, key risks, opportunities, and summary

⚙️ Persistent Settings

  • API key, base URL, model name — all saved locally via Tauri Plugin Store
  • Settings survive app restarts; switching between providers remembers your config

🏗️ Architecture

Frontend (React/Vite)
    ↓ Tauri IPC (invoke)
Tauri Core (Rust) — reads config, spawns sidecar
    ↓ CLI args + stdout
Sidecar (Bun) — Playwright scraping + AI analysis

Three isolated layers with strict unidirectional dependencies. The Rust layer is a thin, tested orchestrator; the Bun sidecar handles all I/O-heavy work.

📦 Installation

macOS (Apple Silicon)

  1. Download StockAI_0.1.0_aarch64.dmg
  2. Open the .dmg and drag StockAI to your Applications folder
  3. If Gatekeeper blocks it: Right-click the app → Open → Open anyway
    (This is expected for unsigned builds — the app contains no network backdoors)

Windows

  1. Download StockAI_0.1.0_x64_en-US.msi (recommended) or the .exe installer
  2. Run the installer and follow the prompts
  3. If SmartScreen warns you, click "More info" → "Run anyway"

Linux (Ubuntu / Debian)

  1. Download StockAI_0.1.0_amd64.deb
  2. Install with: 
    sudo dpkg -i StockAI_0.1.0_amd64.deb
  3. Launch from your app menu or run stockai in the terminal

Note: Linux builds require the WebKitGTK runtime (usually pre-installed on GNOME-based distros).

⚡ Quick Start

  1. Open StockAI and click the ⚙️ Settings icon
  2. Choose your AI provider:
    • OpenAI: Enter your API key and (optionally) a custom base URL
    • Ollama: Make sure Ollama is running locally (ollama serve) and select your model
  3. Go back to the main screen, type a stock ticker (e.g. AAPL, TSLA, NVDA)
  4. Click Analyze and watch the pipeline execute in real time

🛠️ Tech Stack

Layer Technology
UI React 18 + TypeScript + Vite + Tailwind CSS
Desktop shell Tauri v2 (Rust)
Scraping Bun + Playwright (headless Chromium)
AI inference OpenAI SDK (works with any OpenAI-compatible endpoint)
Local AI Ollama (default: qwen3.5:27b)
Packaging Tauri bundler → DMG, MSI/EXE, DEB
CI/CD GitHub Actions (matrix: macOS ARM, Windows, Ubuntu 24.04)

🗺️ Roadmap

  • Multi-stock watchlist — analyze a portfolio in one click
  • Historical trend charts — sentiment over time, powered by cached results
  • Custom prompt templates — tailor the analysis style to your needs
  • Export to PDF/Markdown — shareable analysis reports
  • Intel Mac support — once a signing workflow is in place
  • Streaming output — see AI tokens as they arrive

🤝 Contributing

PRs and issues are very welcome! See CONTRIBUTING guidelines or just open an issue.

If StockAI saves you research time, consider giving it a ⭐ — it helps the project grow!

Built with Tauri, Bun, React, and Ollama.

Assets 7
Loading