On Running MCP

What is On Running MCP?

On Running MCP is a Model Context Protocol server that provides an interface to the On Running website’s product data through a FastMCP server. It allows AI agents to query products, retrieve names, images, and activities, and filter by type, subtype, and gender.

How to use On Running MCP?

Installation requires Python 3.12+ and uv. Clone the repository, create a virtual environment with uv venv, then run uv sync. Add the server to Claude Desktop by running uv run mcp install server.py from the on_running_mcp directory. Use the available tools to query and filter product data.

Key features of On Running MCP

• Query the On Running API with optional filters.
• Get all product names from query results.
• Retrieve activities associated with each product.
• Filter products by name with filter_by_product_name.
• Get an image for a specific product.
• List product types, subtypes, and gender options.

Use cases of On Running MCP

• Find running shoes or apparel for a specific event (e.g., marathon socks).
• Browse products filtered by gender (men's, women's, kids).
• Retrieve product images for visual recommendations.
• List available product types and subtypes to guide queries.
• Narrow down products by name for precise matches.

FAQ from On Running MCP

What does On Running MCP do?

It acts as a bridge between AI assistants and the On Running website API, enabling natural‑language queries to fetch product data with filters.

What are the system requirements?

You need Python 3.12 or newer and the uv package manager.

How do I install On Running MCP?

Clone the repository, create a virtual environment with uv venv, run uv sync to install dependencies, then execute uv run mcp install server.py to register the server with Claude Desktop.

What filters are available?

You can filter by product type (shoes, apparel, accessories), subtype (tops, shorts, etc.), and gender (men’s, women’s, kids). Use the corresponding tool options to see valid values.

How can I get product images?

Use the get_product_image tool to retrieve an image for a specific product after querying the API.