Get Started with
MCP Analytics

Connect Claude to professional statistical analysis tools in 3 simple steps. Instant access with Auth0 single sign-on - no email verification needed.

1

Install MCP Analytics - Choose Your Method

One-click installation or manual configuration - you choose!

📦 Prerequisites

Before installing MCP Analytics, ensure you have:

  • Node.js and npm installed - Required for the npx command
  • Node.js version 5.2 or higher (includes npx)

Download Node.js → or check if installed: node --version

🚀 Option A: One-Click Installation (Recommended)

Claude Desktop Extension (.mcpb)

  1. Download from mcpanalytics.ai/install
  2. Double-click the .mcpb file
  3. Click "Install" in Claude Desktop
  4. Complete Auth0 authentication

Platform-Specific Installers

  • 🪟 Windows: PowerShell installer
  • 🍎 macOS: Shell script installer
  • 🐧 Linux: Bash installer
  • 📝 VS Code: Marketplace extension

Visit install page for all options

⚙️ Option B: Manual Configuration

For advanced users - add to Claude Desktop config:

Location: Settings → Developer → Edit Config

claude_desktop_config.json
// Manual config option - or use one-click installer above! { "mcpServers": { "mcpanalytics": { "command": "npx", "args": [ "-y", "mcp-remote@latest", "https://api.mcpanalytics.ai/auth0" ] } } }
✓ Configuration ready - save and restart Claude Desktop
2

Authenticate with Auth0

Complete the secure Auth0 login in your browser

What happens:

When you restart Claude Desktop, Auth0 will open in your browser for secure authentication:

  • Browser opens automatically - Auth0 secure login page
  • Sign in or create account - use Google, GitHub, or email
  • Instant activation - no email verification needed
  • MCP connection established - tools available immediately
  • Tokens managed automatically - secure OAuth2 flow

Check connection status:

Type /mcp in Claude to see server status and available tools

Claude Desktop
⏺ Checking MCP server status...
mcpanalytics - about
(MCP)(topic: "platform")
{
"company": "PeopleDrivenAI",
"description": "Enterprise AI-powered statistical analysis platform",
... +48 lines (ctrl+r to expand)
}
mcpanalytics - discover.tools
(MCP)(query: "list available tools", limit: 10)
[
"linear_regression",
"ridge_regression",
"lasso_regression",
"arima",
"correlation",
"t_test",
"anova",
"chi_square",
"pca",
"time_series_forecast"
]
✓ Comprehensive statistical toolkit available via MCP
3

Run Your First Analysis

Start analyzing data with natural language requests

How it works:

With Auth0 authentication complete, you have 24-hour demo access with 100 free analyses. You can:

📦 Option 1: Use Test Datasets (Quickest)

Pre-loaded datasets for immediate testing:

  • Customer 360 Dataset - uuid://ds_ds_customers_20250918_1d6p (10,000 rows)
  • Public URLs - Direct analysis of CSV/JSON from GitHub
📤 Option 2: Upload Your Own Data

Secure upload with encryption:

  • Generate upload token with datasets.upload
  • Upload via cURL with your secret key
  • Access with uuid://dataset_id:your_secret

Example walkthrough:

Let's see both options in action - uploading your own data and using test datasets.

▶ You
I need to analyze some data. Can you show me how to upload my own dataset and also what test datasets are available?
⚬ Claude
I'll show you both options. First, let me list the available test datasets, then I'll generate an upload token for your own data.
⏺ Claude
mcpanalytics - datasets.list
(MCP)(category: "testing", limit: 3)
📦 Test Datasets Available:
1. Customer 360 Test Dataset
UUID: ds_ds_customers_20250918_1d6p
Key: kmuKSZ85hnM7stlkA2N6Yaff0aQ9GYR9hezPMAY18uA
✓ Ready to use - no upload needed!
2. Sales Performance Dataset
UUID: ds_sales_20250905_pyvq
Key: 8NZqeHJZtH96sPIV6prNTCw-4Ni1Pf-iDpNP6u5bnpE
✓ Ready to use - no upload needed!
⏺ Claude
mcpanalytics - datasets.upload
(MCP)(expires_in: 300, prefix: "mydata")
✅ Upload token generated
📤 Upload Your Dataset:
Run this command:
curl -X POST https://api.mcpanalytics.ai/datasets/upload/file \
-F "token=upt_P3zCgEEPR8_xRAJ7KvC5u2X02..." \
-F "file=@your_data.csv" \
-F "secret=your-password"
After upload, use: "uuid://ds_mydata_20250908_abc3:your-password"
▶ You
Great! Let's use the Customer 360 test dataset to predict customer churn.
⏺ Claude
mcpanalytics - tools_run
(MCP)(
tool_name: "analytics__statistical__classification__logistic",
taskList: {
"inputs": {
"dataset": "uuid://ds_ds_customers_20250918_1d6p:kmuKSZ85hnM7stlkA2N6Yaff0aQ9GYR9hezPMAY18uA",
"userContext": {"company": "E-commerce", "objective": "Predict customer churn"},
"target": "churn",
"features": ["tenure", "monthly_charges", "total_charges"]
}
},
generate_insights: true
)
✅ SUCCESS: tools_run
📊 Interactive Visual Report Available!
https://api.mcpanalytics.ai/rpt/rpt_CustomerChurn_Analysis
Model Results:
• Model Quality: Good
• R² Score: 0.641
• RMSE: 4.92
• Significant Predictors: 3
... +300 lines available in report
⚬ Claude
Perfect! The analysis is complete. Here's your interactive report: Key findings from the Boston Housing analysis:
• Model Quality: R² = 0.641 (Good - 64.1% variance explained)
• Room Effect: Each additional room increases price by $3,800
• Lower Status %: Strong negative correlation with housing prices
• Student-Teacher Ratio: Lower ratios correlate with higher prices

💡 Quick Tips:
• Use test datasets for quick experimentation (no upload needed)
• Use public URLs for GitHub CSV files (direct analysis)
• Use datasets.upload for your private data (encrypted storage)

🎉 You're ready to go!

You have 24-hour demo access with 100 free analyses. Explore what MCP Analytics can do:

Explore 10 Tools View Pricing Visit Dashboard