home / mcp / performance mcp server
Provides a focused analysis of mobile performance data to detect severe issues via HTTP or local stdio interfaces.
Configuration
View docs{
"mcpServers": {
"dasheng1994-mcp_analyze_quality": {
"url": "http://localhost:8000/meminfo.csv"
}
}
}You deploy and run a lightweight MCP server that analyzes mobile app performance data to detect severe issues. It exposes a simple interface for remote or local analysis and keeps output focused on critical warnings to help you act quickly.
Use the server from your MCP client to analyze performance data streams or files. You can access the live data via an HTTP endpoint or run the analysis locally on your own machine or a dedicated server.
To analyze data from a remote or local source, point your MCP client to the available entry points. If you have a local HTTP server running on port 8000, you can fetch the sample data file meminfo.csv at the following URL and submit it to the MCP client for analysis.
Prerequisites: you need Git and Python 3 on your workstation or server.
git clone [email protected]:DaSheng1994/mcp_analyze_quality.git
cd mcp_analyze_quality
# Create and activate a Python virtual environment
python3 -m venv .venv
source .venv/bin/activate # Linux/macOS
# Install dependencies
pip install -r requirements.txtConfigure the local MCP server for your environment by pointing to the Python interpreter and the main script inside your project. This example uses a virtual environment path as shown in the configuration snippet.
# Local stdio MCP command configuration (example)
"mcpServers": {
"performance_analyzer": {
"type": "stdio",
"command": "/path/to/your/project/.venv/bin/python",
"args": ["/path/to/your/project/main.py"],
"cwd": "/path/to/your/project"
}
}For remote deployment, install and run the server on a host. The following steps show a typical deployment pattern with a background process and a simple HTTP server for data access.
# Server-side deployment on a host
git clone [email protected]:DaSheng1994/mcp_analyze_quality.git
cd mcp_analyze_quality
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
# Start the analysis service in the background
nohup python main.py > mcp.log 2>&1 &
nohup python3 -m http.server 8000 > http.log 2>&1 &If you want team members to run the server from their own environment, provide a client configuration that points to the remote host. The following example shows how to set up the client to SSH into the server and run the analysis script.
{
"mcpServers": {
"performance_analyzer": {
"command": "ssh",
"args": ["your-server", "cd /path/to/mcp_analyze_quality && .venv/bin/python main.py"],
"env": {}
}
}
}Usage notes: The server is designed to produce concise output that highlights only severe warnings. The current severity checks include a physical memory threshold and a rapid increase in the number of views. You can customize rules by editing the provided rules file in your Cursor setup to tailor the analysis to your needs.
Security and environment: If you deploy remotely, ensure access is restricted to authorized clients. Use network security best practices and consider placing the HTTP data endpoint behind authentication if exposing meminfo.csv externally.
Troubleshooting tips: If the server fails to start, verify that the Python virtual environment is activated, required dependencies are installed, and that the port 8000 is available for the HTTP server. Check logs in mcp.log and http.log for clues.
Rule customization: You can modify the analysis rules to fit your data sources. The rules file for quality analysis is located under your Cursor rules directory and can be edited to adjust thresholds or add new checks.