A self-hosted price tracking application, PriceGhost monitors product prices across any website, offering users precise control via a multi-strategy extraction system and optional AI enhancements. It targets users seeking accurate price monitoring, timely notifications for drops or restocks, and a self-hosted solution, providing a robust alternative to commercial trackers.

How It Works

PriceGhost employs a novel multi-strategy price extraction system, combining JSON-LD, site-specific scrapers, generic CSS selectors, and AI analysis (Claude/GPT/Ollama) in parallel. Its core differentiator is the "Price Voting Modal," which presents price candidates from each method, allowing users to manually select the correct price. This approach ensures accuracy, preventing common tracking errors like mistaking discount amounts for actual prices. AI features are strongly recommended for enhanced compatibility and as a fallback for difficult-to-scrape sites.

Quick Start & Requirements

• Primary Install: Docker is recommended. Clone the repository, navigate to the directory, and run docker-compose up -d.
• Prerequisites: Node.js 18+, PostgreSQL 14+. Optional AI providers (Anthropic, OpenAI, Ollama) require API keys or local setup.
• Environment Variables: Key variables include POSTGRES_USER, POSTGRES_PASSWORD, JWT_SECRET, and DATABASE_URL.
• Links: Official GitHub repository (implied).

Highlighted Details

• Multi-Strategy Extraction: Leverages JSON-LD, site-specific scrapers, CSS, and AI with a user-driven price selection modal.
• AI Integration: Optional AI (Claude, GPT, Ollama) for fallback, verification, and arbitration, improving accuracy on complex e-commerce sites.
• Extensive Notifications: Supports Telegram, Discord, Pushover, ntfy.sh, and Gotify for alerts.
• PWA Support: Enables installation on mobile devices via "Add to Home Screen."
• Headless Browser: Utilizes Puppeteer with stealth for JavaScript-heavy sites.

Maintenance & Community

The README does not detail specific contributors, sponsorships, or community channels like Discord or Slack. The project highlights its development using Claude Code.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license permits commercial use and integration with closed-source projects without significant restrictions.

Limitations & Caveats

Enabling advanced AI features requires API keys, potentially incurring costs, or local compute resources. While robust, aggressive scraping intervals or highly sophisticated anti-scraping measures on some websites may still present challenges. The project relies on external AI providers or local Ollama setup for its most advanced capabilities.