JunoJunHyun Festival Finder MCP

What JunoJunHyun-festival-finder-mcp MCP server does

JunoJunHyun-festival-finder-mcp MCP server provides access to performance and festival data from KOPIS, Korea’s integrated performing arts information network. The repository separates the event-data logic from the code that exposes it to other platforms. This makes the retrieval and processing layer reusable without tying it to a single client or channel.

The available project adapters target two connection patterns. kakao_server.py is intended for Kakao PlayMCP integration, while web_server.py exposes a general web API for browsers or other applications. The supplied material describes Korean performance and festival data; it does not establish coverage of worldwide events.

How it works

The core engine in core_logic.py calls the KOPIS API and returns processed data as standard Python values. It does not handle the details of the platform consuming that data. Instead, each adapter receives requests in its own format, invokes the core logic, and formats the response for its target platform.

For web testing, the README gives an endpoint pattern based on http://localhost:5001/api/performances?stdate=.... The Kakao path runs on port 5000 and is designed to be exposed through an ngrok tunnel, with the resulting /kakao URL registered as a PlayMCP endpoint. These are HTTP-style adapter servers rather than a documented local stdio package command.

Setup and configuration

Install the dependencies listed in requirements.txt with pip. Before starting an adapter, set the KOPIS_API_KEY variable in core_logic.py to an API key issued to you. The provided setup instructions do not describe an environment-variable-based configuration flow, so the key is configured in the source file as documented.

Start python kakao_server.py when testing the Kakao integration. Connect ngrok to http://localhost:5000, then register the resulting address with the Kakao PlayMCP endpoint configuration. For a general web API test, start python web_server.py and send requests to the local API on port 5001.

Tools and capabilities

JunoJunHyun-festival-finder-mcp MCP server supports:

• Retrieving performance and festival information from KOPIS.
• Processing API results through platform-independent Python logic.
• Serving Kakao PlayMCP requests through the Kakao adapter.
• Providing a general web API through the web adapter.
• Accepting performance queries using parameters such as stdate in the documented web endpoint pattern.

Limitations and notes

The repository requires a KOPIS API key and does not document anonymous access. The README does not provide a named MCP tool inventory, schema reference, response examples, or filtering details beyond the shown stdate parameter. It also does not identify a license or list direct compatibility with Claude Desktop, Cursor, Windsurf, or Cline.

The Smithery command shown in the README is an external installation route, not the project’s own package command. The documented project startup method runs Python adapter files directly, and the supplied material does not specify a stdio installation command or a hosted remote endpoint.