PanWatch is an AI-driven, self-hosted stock monitoring assistant designed for individual investors and traders. It provides real-time market data, intelligent technical analysis, and multi-account portfolio management, offering a private and AI-native approach to investment decision support. The primary benefit is gaining personalized, AI-powered insights without compromising data privacy through third-party services.

How It Works

PanWatch employs an AI Agent system that analyzes market data and user holdings for pre-market strategy generation, intra-day anomaly detection, post-market review, and relevant news filtering. Its core advantage lies in its "AI Native" design, moving beyond simple indicator stacking to enable AI to understand user portfolios, trading styles, and objectives. The self-hosted architecture ensures data privacy, as holdings data remains under user control.

Quick Start & Requirements

• Primary install: Docker is the recommended deployment method. A docker run command and docker-compose example are provided for quick setup.
• Prerequisites: Docker is essential. For local development, Python 3.10+, Node.js 18+, and pnpm are required. The Docker image includes Playwright, which automatically downloads Chromium on first container start if browser capabilities are needed.
• Setup: Docker deployment is estimated at 5 minutes for configuration, though the initial Chromium download may add a few minutes.
• Links: Web UI: http://localhost:8000. Community: t.me/panwatch.

Highlighted Details

• AI Agent System: Features automated pre-market analysis, real-time intra-day monitoring for signals (e.g., RSI/KDJ/MACD resonance), post-market daily reports, and AI-filtered financial news relevant to holdings.
• Comprehensive Technical Analysis: Includes trend indicators (MA, MACD, Bollinger Bands), momentum indicators (RSI, KDJ), volume-price analysis, candlestick pattern recognition, and automatic support/resistance level calculation.
• Multi-Market & Multi-Account: Supports real-time data for A-shares, Hong Kong stocks, and US stocks, with independent management for multiple brokerage accounts and consolidated asset display.
• Flexible Price Alerts: Customizable alerts based on price, percentage change, volume, etc., with support for various notification channels including Telegram, WeChat, DingTalk, Lark, Bark, and custom webhooks.

Maintenance & Community

The project is actively maintained with a GitHub Actions CI/CD pipeline for Docker image releases. Community interaction and support are primarily facilitated through a Telegram group. Contributions via Issues and Pull Requests are welcomed, with a dedicated contribution guide available.

Licensing & Compatibility

PanWatch is released under the MIT License. This permissive license allows for broad use, modification, and distribution, including for commercial purposes and integration within closed-source applications without significant restrictions.

Limitations & Caveats

The AI features require configuration with an OpenAI-compatible API endpoint. Self-hosting necessitates managing Docker and its dependencies. While the Docker image handles Chromium installation, it requires network access and may take time on the first run.