Skip to main content
Glama
wonyoungseong

GA4 MCP Server

README.md
# GA4 MCP Server

A Model Context Protocol (MCP) server for Google Analytics 4 (GA4) that provides tools for querying GA4 data through the Admin API and Data API.

## Features

- **Dual Authentication**: Supports both OAuth Playground tokens and Service Account credentials
- **8 GA4 Tools**:
  - Account and property management (Admin API)
  - Report execution and realtime data (Data API)
  - **GTM → GA4 parameter validation** (cross-platform orchestration)
- **Compatible with Claude Desktop** and other MCP-enabled clients
- **GTM Integration**: Works with GTM MCP Server for end-to-end event validation

## Installation

```bash
npm install
npm run build
```

## Quick Start (OAuth Setup)

### 1. Create OAuth Credentials

1. Go to [Google Cloud Console](https://console.cloud.google.com/apis/credentials)
2. Create a new project or select an existing one
3. Enable the **Google Analytics Data API** and **Google Analytics Admin API**
4. Go to **Credentials** → **Create Credentials** → **OAuth client ID**
5. Select **Desktop app** as the application type
6. Download or copy the Client ID and Client Secret

### 2. Run Setup Script

```bash
# Set your OAuth credentials
export GA4_CLIENT_ID="your-client-id.apps.googleusercontent.com"
export GA4_CLIENT_SECRET="your-client-secret"

# Run setup to authenticate with Google
npm run setup
```

This will:
- Open a browser for Google login
- Request read-only access to Google Analytics
- Save tokens to `~/.ga4-mcp/tokens.json`

### 3. Use the Server

Once authenticated, you can use the MCP server with Claude Desktop or other clients.

## Authentication

The server supports two authentication methods, checked in this order:

### 1. OAuth Playground Tokens (Recommended for personal use)

Set these environment variables:
- `GA4_ACCESS_TOKEN` - OAuth access token
- `GA4_REFRESH_TOKEN` - OAuth refresh token
- `GA4_CLIENT_ID` - OAuth client ID
- `GA4_CLIENT_SECRET` - OAuth client secret

Or create `~/.ga4-mcp/tokens.json`:
```json
{
  "access_token": "your-access-token",
  "refresh_token": "your-refresh-token",
  "client_id": "your-client-id",
  "client_secret": "your-client-secret"
}
```

### 2. Service Account (Recommended for production)

Options (in priority order):
1. Set `GA4_SERVICE_ACCOUNT_JSON` environment variable with JSON string
2. Set `GOOGLE_APPLICATION_CREDENTIALS` to the JSON file path
3. Place JSON file at `~/.ga4-mcp/credentials.json`
4. Place JSON file in `./Credential/` folder

## Usage

### Claude Desktop Configuration

Add to your Claude Desktop configuration (`~/.config/claude-desktop/config.json`):

```json
{
  "mcpServers": {
    "ga4": {
      "command": "node",
      "args": ["/path/to/ga4-mcp-server/dist/index.js"],
      "env": {
        "GA4_ACCESS_TOKEN": "your-access-token",
        "GA4_REFRESH_TOKEN": "your-refresh-token",
        "GA4_CLIENT_ID": "your-client-id",
        "GA4_CLIENT_SECRET": "your-client-secret"
      }
    }
  }
}
```

### Direct Execution

```bash
# With OAuth environment variables
GA4_ACCESS_TOKEN="..." GA4_REFRESH_TOKEN="..." GA4_CLIENT_ID="..." GA4_CLIENT_SECRET="..." node dist/index.js

# With Service Account
GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json" node dist/index.js
```

## Available Tools

### Admin API Tools

| Tool | Description |
|------|-------------|
| `ga4_account_summaries` | List all GA4 accounts and properties the user has access to |
| `ga4_property_details` | Get detailed information about a specific property |
| `ga4_google_ads_links` | List Google Ads accounts linked to a property |
| `ga4_property_annotations` | List annotations for a property (limited support) |

### Data API Tools

| Tool | Description |
|------|-------------|
| `ga4_run_report` | Run a standard GA4 report with dimensions, metrics, and date ranges |
| `ga4_run_realtime_report` | Run a realtime report for the last 30 minutes |
| `ga4_custom_dimensions_metrics` | Get custom dimensions and metrics defined for a property |

### GTM Validation Tool

| Tool | Description |
|------|-------------|
| `ga4_validate_gtm_params` | Validate GTM event parameters against GA4 custom dimensions and data collection |

## Example Usage

### Get Account Summaries
```json
{
  "tool": "ga4_account_summaries"
}
```

### Run a Report
```json
{
  "tool": "ga4_run_report",
  "arguments": {
    "propertyId": "123456789",
    "dateRanges": [
      {"startDate": "30daysAgo", "endDate": "yesterday"}
    ],
    "dimensions": ["country", "deviceCategory"],
    "metrics": ["activeUsers", "sessions"]
  }
}
```

### Run Realtime Report
```json
{
  "tool": "ga4_run_realtime_report",
  "arguments": {
    "propertyId": "123456789",
    "dimensions": ["country"],
    "metrics": ["activeUsers"]
  }
}
```

### Validate GTM Parameters
```json
{
  "tool": "ga4_validate_gtm_params",
  "arguments": {
    "propertyId": "123456789",
    "gtmEvents": [
      {
        "eventName": "purchase",
        "parameters": ["transaction_id", "value", "currency"]
      }
    ],
    "startDate": "7daysAgo",
    "endDate": "yesterday"
  }
}
```

Or with GTM Export JSON:
```json
{
  "tool": "ga4_validate_gtm_params",
  "arguments": {
    "propertyId": "123456789",
    "gtmExportJson": { "...GTM container export data..." }
  }
}
```

## GTM Integration (Claude Agent Orchestration)

This server works with the GTM MCP Server to provide end-to-end validation of GTM event parameters in GA4.

### Architecture

```
┌─────────────────────────────────────────────────────────────────────┐
│                        Claude Agent                                 │
│                    (Orchestrator Role)                              │
├─────────────────────────────────────────────────────────────────────┤
│                                                                     │
│  ① GTM MCP Server                  ② GA4 MCP Server                │
│  (gtmAgent/mcp-server)             (ga4-mcp-server)                 │
│                                                                     │
│  gtm_export_full()                 ga4_validate_gtm_params()        │
│  gtm_tag().list()                  ga4_custom_dimensions_metrics()  │
│       │                            ga4_run_report()                 │
│       │                                   ▲                         │
│       └───────────────────────────────────┘                         │
│                Agent relays data                                    │
│                                                                     │
└─────────────────────────────────────────────────────────────────────┘
```

### Workflow

1. **Extract from GTM**: `gtm_export_full()` → Get GA4 Event tags and parameters
2. **Validate in GA4**: `ga4_validate_gtm_params()` → Check registration and collection
3. **Get recommendations**: Parameters not registered or not collecting data

### Claude Desktop Configuration (Both Servers)

```json
{
  "mcpServers": {
    "gtm": {
      "command": "node",
      "args": ["/path/to/gtmAgent/mcp-server/dist/index.js"],
      "env": {}
    },
    "ga4": {
      "command": "node",
      "args": ["/path/to/ga4-mcp-server/dist/index.js"],
      "env": {
        "GA4_ACCESS_TOKEN": "your-access-token",
        "GA4_REFRESH_TOKEN": "your-refresh-token",
        "GA4_CLIENT_ID": "your-client-id",
        "GA4_CLIENT_SECRET": "your-client-secret"
      }
    }
  }
}
```

### API Call Efficiency

| Scenario | Traditional | Optimized | Reduction |
|----------|-------------|-----------|-----------|
| 5 events × 21 params | 105 calls | 22 calls | 79% |
| 14 events × 32 params | 448 calls | 33 calls | 93% |

The validation tool queries metadata once, then makes one API call per unique parameter instead of per event-parameter combination.

## API Documentation References

- [GA4 Admin API](https://developers.google.com/analytics/devguides/config/admin/v1)
- [GA4 Data API](https://developers.google.com/analytics/devguides/reporting/data/v1)
- [Standard Dimensions](https://developers.google.com/analytics/devguides/reporting/data/v1/api-schema#dimensions)
- [Standard Metrics](https://developers.google.com/analytics/devguides/reporting/data/v1/api-schema#metrics)
- [Realtime Dimensions](https://developers.google.com/analytics/devguides/reporting/data/v1/realtime-api-schema#dimensions)
- [Realtime Metrics](https://developers.google.com/analytics/devguides/reporting/data/v1/realtime-api-schema#metrics)

## Required Google API Scopes

The server uses the following scope:
- `https://www.googleapis.com/auth/analytics.readonly`

## License

MIT

TDQS

A3.9/5.0

Scored across 8 tools

Disambiguation5/5

Each tool has a clearly distinct purpose targeting specific GA4 resources or operations. For example, ga4_account_summaries retrieves account-level metadata, ga4_run_report handles standard analytics queries, and ga4_validate_gtm_params focuses on data quality validation. There is no functional overlap between tools, making tool selection straightforward for an agent.

Naming Consistency5/5

All tools follow a consistent 'ga4_' prefix with descriptive snake_case naming that clearly indicates their function (e.g., ga4_property_details, ga4_run_realtime_report). The naming pattern is uniform throughout the set, making it predictable and easy to understand.

Tool Count5/5

With 8 tools, this server is well-scoped for GA4 operations. It covers essential areas like metadata retrieval (accounts, properties, custom definitions), reporting (standard and realtime), and data validation, without being overwhelming or sparse. Each tool serves a clear, necessary function in the analytics workflow.

Completeness4/5

The tool set provides strong coverage for core GA4 operations including metadata access, reporting, and data validation. However, there are minor gaps such as the absence of tools for creating or managing resources (e.g., creating custom dimensions or annotations) or handling administrative tasks like user permissions, which could limit full lifecycle management.

Maintenance

ActivityInactive
ResponsivenessNo issues