home / mcp / perfdog to tableau mcp servers mcp server

PerfDog to Tableau MCP Servers MCP Server

로그인 , 다운로드 , 태블로 변환

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "kimjeonghyun225-cpu-perfdog_mcp": {
      "command": "python",
      "args": [
        "C:/path/to/perfdog_MCP/perfdog_mcp_server.py"
      ]
    }
  }
}

PerfDog to Tableau MCP Servers automate the download of PerfDog mobile game performance data and convert it into Tableau-ready CSV formats. This setup lets you fetch project data in bulk and produce both summary and per-second time series exports for Tableau dashboards and performance analysis.

How to use

You can run the PerfDog downloader MCP server and the Tableau converter MCP server from your MCP client. The two servers work together to fetch PerfDog project data and transform it into two CSV formats: a Summary CSV with device-level statistics and a TimeSeries CSV with per-second metrics. You can trigger both servers from your client session, provide the PerfDog project URL, choose an output location, and then start the conversion to Tableau-ready files.

How to install

Prerequisites: Python and pip must be available on your system.

pip install git+https://github.com/kimjeonghyun225-cpu/perfdog_MCP.git

Additional setup and configuration

Configure the local Claude Desktop integration to run the two MCP servers. The following example shows two entries you add to Claude desktop configuration under mcpServers.

{
  "mcpServers": {
    "perfdog": {
      "command": "python",
      "args": ["C:/path/to/perfdog_MCP/perfdog_mcp_server.py"]
    },
    "tableau_converter": {
      "command": "python",
      "args": ["C:/path/to/perfdog_MCP/tableau_converter_MCP.py"]
    }
  }
}

Usage notes

With Claude Skill: install the perfdog-tableau-converter.skill and upload it to Claude.ai. Then you can run the conversion by providing the PerfDog project URL, the desired output path, and the project name. If you prefer direct MCP usage, Claude will prompt you for credentials, project URL, output path, and project name, then download the Excel files and convert them to Tableau CSV formats before reporting completion.

Output files

Summary CSV: summary_{project}.csv contains device-level statistics such as device type, platform, FPS metrics, CPU/Memory, jank indicators, and stability metrics.

TimeSeries CSV: fps_{project}.csv contains per-second measurements including FPS, CPU, Memory, GPU, network, temperature, and platform-specific metrics.

Configuration

Session files and logs are stored in the following locations for troubleshooting and audit purposes.

Session: ~/.perfdog_mcp/perfdog_session.json Logs: ~/.perfdog_mcp/perfdog_downloader.log and tableau_converter.log Downloads: ~/Desktop/perfdog_downloads/

Troubleshooting

MCP server not recognized: ensure Claude Desktop is fully restarted and paths in the config are correct. Confirm Python is accessible from the command line.

Path issues: on Windows use double backslashes, on macOS/Linux use forward slashes.

Support

For issues, open GitHub Issues on the project page or contact the internal KRAFTON Compatibility QA Team.

Notes

This MCP setup is designed for automating PerfDog data gathering and Tableau-ready conversion. It includes two local servers that you run via Python scripts and integrates with Claude Desktop for seamless operation.

Available tools

PerfDog downloader MCP server

Downloads PerfDog Excel files in bulk for specified projects and interfaces with case metadata APIs to enrich data.

Tableau converter MCP server

Converts downloaded PerfDog data into two Tableau-friendly CSV formats: a summary CSV and a per-second time series CSV.