Marketfiyati

You deploy the Marketfiyati MCP Server to access Turkish market product prices from marketfiyati.org.tr through a standardized MCP API. It enables fast product searches, location-based queries, and multi-market price comparisons, making it easier to integrate price data into AI tools and clients.

How to use

You connect to the MCP server from your MCP client by pointing it at either a local stdio instance or a remote HTTP endpoint. The server exposes endpoints for searching products, checking health, and retrieving the MCP interface for AI tool integration. Use the provided paths to perform searches by keywords, categories, and location, and to access the MCP protocol endpoint for seamless tool integration.

How to install

Prerequisites you need before running the server are Python 3.x and a working internet connection. You will also need Docker if you choose the containerized route.

Option A: Install with Docker (recommended) setup

# Pull the pre-built image
docker pull ghcr.io/yibudak/marketfiyati_mcp:latest

# Run the container
docker run -d -p 8000:8000 --name marketfiyati-mcp ghcr.io/yibudak/marketfiyati_mcp:latest

# Run with SOCKS proxy (optional)
docker run -d -p 8000:8000 -e SOCKS_PROXY=socks5://localhost:1080 --name marketfiyati-mcp ghcr.io/yibudak/marketfiyati_mcp:latest

How to run locally (manual installation)

If you prefer running the server directly on your machine, install the Python dependencies and start the server with Uvicorn.

# Clone the project repository
git clone https://github.com/yibudak/marketfiyati_mcp.git
cd marketfiyati_mcp

# Install dependencies
pip install -r requirements.txt

# Run the server
uvicorn app.main:app --port 8000

# Run with SOCKS proxy (optional)
SOCKS_PROXY=socks5://localhost:1080 uvicorn app.main:app --port 8000

Additional setup and usage notes

SOCKS proxy support can route all external calls to marketfiyati.org.tr through a proxy. You can provide the proxy via environment variable SOCKS_PROXY before starting the server.

For easy integration with Claude Desktop, you can configure the MCP endpoint like this: a local MCP URL under the marketfiyati key.

API and endpoints you’ll use

The server exposes an MCP endpoint at /mcp for AI tooling integration, and also provides health and version checks as /health and /version. You can explore the interactive API docs at /docs once the server is running.

Example client configuration for Claude Desktop

json
{
  "mcpServers": {
    "marketfiyati": {
      "url": "http://localhost:8000/mcp"
    }
  }
}

Available tools

/mcp endpoint

HTTP MCP endpoint exposed at /mcp for AI tool integration.

/health

Health check endpoint that reports server status and version.

/version

Endpoint to retrieve the API version.

/docs

Interactive API documentation (Swagger UI) available after startup.