PriceGhost is a self-hosted price tracking application for monitoring product prices on e-commerce sites. The project describes itself as a way to track prices, receive alerts when prices drop or reach a target, and notice when an item comes back in stock.
The repository presents PriceGhost as a full application rather than a narrow script: it includes a frontend, backend, database setup, Docker Compose file, notification settings, user accounts, and price history views. Its strongest editorial angle is control. Instead of treating price extraction as a black box, PriceGhost exposes multiple detected price candidates when methods disagree, so the user can decide which value should be tracked.
Why this kind of tool matters
Price tracking looks simple until a shop page contains several competing numbers: list prices, discount amounts, bundles, financing plans, shipping adjustments, membership prices, and temporary promotions. A tracker that silently chooses the wrong number can generate misleading alerts or build an inaccurate history.
PriceGhost tries to address that problem by making price extraction more transparent. Its README describes a multi-strategy approach in which several methods can identify candidate prices, with a selection flow for cases where the application is not confident enough to pick one automatically. For people who track expensive items or compare several retailers over time, that decision point can be more valuable than a fully automatic but opaque result.
What PriceGhost offers
The source material describes PriceGhost as combining several practical layers:
- product price monitoring with configurable checks
- target price, price-drop, and back-in-stock alerts
- price history charts and dashboard indicators
- notification channels including Telegram, Discord, Pushover, ntfy.sh, and Gotify
- multi-user support with account and admin controls
- optional AI-assisted extraction, verification, and arbitration
- Docker-based setup for running the application with its supporting services
The project also lists site-specific scrapers for several major retailers, while falling back to more generic extraction methods and optional AI support for other pages. That mix suggests a tool designed for broad coverage, but still dependent on the structure and behavior of the sites being tracked.
Best-fit scenarios
PriceGhost is best suited to users who want ownership of their tracking data and are comfortable running a small web application. It makes particular sense for homelab users, VPS users, small teams, families, or deal hunters who want one private place to collect products, monitor price history, and route alerts through channels they already use.
It may also appeal to people who dislike relying on retailer-specific trackers. A self-hosted application can be more flexible, especially when a product is spread across several sites or when the user wants to preserve a long-term price record outside a retailer’s own ecosystem.
Adoption notes
The README presents Docker as the recommended way to start PriceGhost, with PostgreSQL as the database and a Node.js/Express backend paired with a React and TypeScript frontend. That is a familiar stack for self-hosters, but it still means the operator is responsible for deployment, updates, secrets, backups, and network exposure.
The notification setup is an important early configuration step. PriceGhost is most useful when alerts are routed somewhere the user will actually notice them, so choosing between Telegram, Discord, Pushover, ntfy.sh, or Gotify should be treated as part of the initial setup rather than an afterthought.
Optional AI features are positioned as helpful for difficult product pages. That can improve extraction in some cases, but it also introduces either API keys and usage costs or local model hosting requirements, depending on the provider chosen.
Caveats and limits
PriceGhost depends on extracting information from pages it does not control. Retailers can redesign layouts, change structured data, add anti-bot measures, vary prices by location, or show different prices to different sessions. The presence of multiple extraction strategies reduces some fragility, but it does not remove the operational reality of web scraping.
Self-hosting also shifts responsibility to the user. A hosted price tracker may hide infrastructure details; PriceGhost asks the user to operate the application, maintain the database, secure access, and understand when a price result looks wrong. The price selection modal is useful precisely because the software cannot always know which number is correct.
Editorial verdict
PriceGhost is an interesting self-hosted alternative to conventional price trackers because it focuses not only on alerts, but also on the trustworthiness of the tracked number. The multi-strategy extraction flow and user-confirmed price selection make it more transparent than a simple scraper that saves the first matching value.
The trade-off is complexity. This is a tool for users who are comfortable managing a Docker-backed application and who understand that e-commerce scraping can be imperfect. For that audience, PriceGhost looks like a practical project with a clear problem focus: track the price you actually meant to track, keep the history yourself, and send alerts through your own preferred channels.
Primary link
Learn more at: https://github.com/clucraft/PriceGhost
