PerfDog to Tableau MCP Server
Automates downloading mobile game performance data from PerfDog and converts it to Tableau-ready CSV format with device statistics, FPS metrics, and time-series data.
README
PerfDog to Tableau MCP Servers
🎮 Automate PerfDog mobile game performance data download and conversion to Tableau-ready CSV format.
🚀 Features
PerfDog Server
- Auto-login with session persistence
- Bulk download Excel performance data from PerfDog projects
- API integration for case metadata
- Smart naming
{Device}_{Case}.xlsx
Tableau Converter
- Summary CSV - Device-level statistics (FPS, CPU, Memory, Jank, etc.)
- TimeSeries CSV - Second-by-second performance metrics
- Platform detection - Auto-detect iOS/Android from device info
- CPU normalization - Proper CPU calculation per platform
- Package name extraction - Include app package in filename
📦 Installation
Quick Install
pip install git+https://github.com/kimjeonghyun225-cpu/perfdog_MCP.git
Manual Setup
-
Clone the repository
git clone https://github.com/kimjeonghyun225-cpu/perfdog_MCP.git cd perfdog_MCP -
Install dependencies
pip install -r requirements.txt -
Configure Claude Desktop
Windows: Edit
%APPDATA%\Claude\claude_desktop_config.jsonmacOS: Edit
~/Library/Application Support/Claude/claude_desktop_config.jsonAdd:
{ "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"] } } } -
Restart Claude Desktop
💡 Usage
With Claude Skill (Recommended)
-
Install the Claude Skill
- Download
perfdog-tableau-converter.skill - Upload to Claude.ai → Settings → Skills
- Download
-
Use with Claude
"Convert this PerfDog project to Tableau: https://perfdog.wetest.net/taskdata/191797/cases Output: C:\Users\me\Desktop\tableau Project name: PUBG_v3.8.0"
Direct MCP Usage
Claude will automatically:
- Ask for PerfDog credentials (first time only)
- Request project URL, output path, and project name
- Download all Excel files
- Convert to Tableau CSV format
- Report completion
📊 Output Files
Summary CSV (summary_{project}.csv)
Device-level statistics including:
- Device info (Type, Product, Platform)
- FPS metrics (Avg, Median, Max, 1% Low)
- Stability indicators (MedRange, Std, Smooth Index)
- CPU & Memory (Avg, Peak)
- Jank metrics (Jank, BigJank, Drop FPS)
TimeSeries CSV (fps_{project}.csv)
Per-second measurements including:
- FPS, CPU, Memory, GPU
- Network (Recv/Send)
- Temperature
- Platform-specific metrics
🛠️ Development
Project Structure
perfdog_MCP/
├── perfdog_mcp_server.py # PerfDog downloader MCP server
├── tableau_converter_MCP.py # Tableau converter MCP server
├── requirements.txt # Python dependencies
├── README.md # This file
└── perfdog-tableau-converter.skill # Claude Skill
Dependencies
mcp>=1.0.0
aiohttp>=3.9.0
aiofiles>=23.2.0
pandas>=2.0.0
openpyxl>=3.1.0
🎯 Use Cases
- QA Performance Testing - Automate test data collection
- Version Comparisons - Compare performance across game versions
- Dashboard Creation - Feed data into Tableau dashboards
- Regression Analysis - Track performance trends over time
🔒 Configuration
Session files and logs are stored in:
- Session:
~/.perfdog_mcp/perfdog_session.json - Logs:
~/.perfdog_mcp/perfdog_downloader.logandtableau_converter.log - Downloads:
~/Desktop/perfdog_downloads/
🐛 Troubleshooting
Login fails
- Verify credentials at https://perfdog.wetest.net
- Check logs in
~/.perfdog_mcp/
MCP server not recognized
- Restart Claude Desktop completely
- Verify paths in
claude_desktop_config.json - Check Python is accessible from command line
Path issues
- Windows: Use double backslashes
C:\\Users\\... - macOS/Linux: Use forward slashes
/Users/...
📞 Support
- GitHub Issues: https://github.com/kimjeonghyun225-cpu/perfdog_MCP/issues
- Internal: KRAFTON Compatibility QA Team
📝 License
Internal use - KRAFTON
🙏 Credits
Developed by KRAFTON Compatibility QA Team for mobile game performance testing automation.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。