Doorzo Deal Monitor
MCP server for finding and monitoring deals on Japanese shopping sites via Doorzo. It enables AI assistants to search across 9 marketplaces, set price alerts, and check for new matches.
README
Doorzo Deal Monitor
A tool that finds good deals on Japanese shopping sites and alerts you when something worth buying shows up.
It searches doorzo.com — a proxy-shopping service that combines 9 Japanese marketplaces: Mercari, Rakuma, PayPay Flea, PayPay Mall, Rakuten, Yahoo Auctions, Amazon Japan, Lashinbang, and SNKRDUNK.
No AI needed to use it. No browser automation or scraping either — it talks to doorzo's own API directly.
What it does
- Search for a keyword (e.g.
マリオパーティornintendo switch) across all 9 marketplaces at once. - Filter by price — set a ceiling in Australian dollars in the web UI.
- Spot discounts — shows the list price vs. the sale price and the % off.
- Monitor keywords — tell it "alert me when a Mario Party game appears for under A$30". It checks in the background and notifies you (macOS notification + a log file).
- Trending searches — see what people are hunting for right now.
- Live AUD prices — the web UI converts Doorzo's JPY prices and price ceilings using Doorzo's current AUD rate.
- Open via Doorzo — product and alert links stay inside Doorzo instead of sending you to the underlying marketplace.
- Choose how to buy — narrow results to Buy It Now listings or auctions.
- Avoid junk listings — hide likely junk, broken, untested, or parts-only items by default; show or isolate them when wanted. Flagged results carry a clear
JUNK / PARTSlabel and reason.
Two ways to use it
| Web UI | AI (MCP) | |
|---|---|---|
| Who uses it | You, in a browser | An AI assistant like Claude Code |
| How it looks | A website at http://127.0.0.1:8756 |
Tools the AI can call |
| Best for | Browsing, adding monitors, checking alerts | Asking the AI to find/watch deals for you |
Both share the same data and the same monitors.
Getting started (3 steps)
1. Install. You need Python 3.11+ and uv (a Python tool installer).
git clone https://github.com/TESTYEE-09/DoorzoMCP.git
cd DoorzoMCP
uv sync
2. Start the web UI.
uv run doorzo-web
3. Open your browser at http://127.0.0.1:8756 — done.
Search something, or add a monitor: give it a name, a keyword, and a max price. Click Check all monitors to check for new matches right now. New matches fire a macOS notification and appear in the Alerts list.
The web UI displays and accepts prices in AUD. Internally, monitor ceilings are converted to JPY so the web UI and MCP server can continue sharing the same backward-compatible state. Product links open through Doorzo, where you can review or purchase the listing using Doorzo's proxy-shopping flow.
Your monitors and alerts are stored in ~/.doorzo-mcp/ and survive restarts.
Set DOORZO_STATE_DIR before starting either service only if you want to keep
that state somewhere else.
Letting an AI use it (optional)
If you use Claude Code (or another MCP-compatible assistant), you can give it access to the same search and monitors:
claude mcp add doorzo-deals -- uv run --directory /path/to/DoorzoMCP doorzo-mcp
The AI then has these tools: doorzo_search, doorzo_hot_searches, doorzo_exchange_rate, doorzo_monitor_add, doorzo_monitor_remove, doorzo_monitor_list, and doorzo_check_monitors. Example: "add a monitor for a Mario Party game under ¥3,000 and check it now" — it will do that and tell you what it found.
The MCP tool parameters and returned price fields remain in JPY. This keeps the public tool contract stable for existing agents; AUD conversion is a web UI feature.
Project layout
doorzo_mcp/
client.py # talks to doorzo.com's API
deals.py # decides what counts as a deal
store.py # saves your monitors and alerts (~/.doorzo-mcp)
notify.py # macOS notifications
server.py # the AI (MCP) tools
web.py # the website and its API
static/ # the web page itself (single file, no build step)
Things to know
- Prices are before proxy fees. You still pay doorzo's service fee and international shipping. The tool finds cheap listings.
- AUD amounts are estimates. Doorzo's live rate is used, and the final amount may differ when Doorzo processes the purchase.
- Condition text comes from the sites in Chinese; this tool translates it to English (unknown labels pass through as-is).
- The API could change. doorzo.com is an unaffiliated third party; this project uses its internal API with no authentication, so it may break if they change things or add bot protection.
- MIT licensed. Use at your own risk. Not affiliated with or endorsed by doorzo.com.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.