Google Analytics 4 MCP Server
Enables managing Google Analytics 4 properties, data streams, conversions, and running reports using natural language through the Admin and Data APIs.
README
Google Analytics 4 MCP Server
<!-- mcp-name: io.github.stucchi/google-analytics -->
MCP server for Google Analytics 4 — create properties, configure data streams, manage conversions, and run reports via the Admin & Data APIs.
Setup (one-time, ~3 minutes)
-
Google Cloud Console → IAM & Admin → Service Accounts
- Create a service account (any name)
- Click it → Keys → Add Key → JSON → Download
-
Enable APIs in your GCP project:
-
GA4 Admin → Property Access Management → Add the service account email (from the JSON) as Editor
-
Configure credentials (pick one):
- Option A (recommended): Copy the JSON content into the
GA_CREDENTIALSenv var in your MCP config - Option B: Save the file to
~/.google-analytics-4-mcp/credentials.json
- Option A (recommended): Copy the JSON content into the
Done. No browser flow, no consent screen, no tokens to refresh.
Installation
Claude Code
claude mcp add google-analytics -e GA_CREDENTIALS='{"type":"service_account",...}' -- uvx google-analytics-4-mcp
Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"google-analytics": {
"command": "uvx",
"args": ["google-analytics-4-mcp"],
"env": {
"GA_CREDENTIALS": "{\"type\":\"service_account\",\"project_id\":\"...\", ...}"
}
}
}
}
From source
git clone https://github.com/stucchi/mcp-google-analytics.git
cd mcp-google-analytics
uv run google-analytics-4-mcp
Getting Started
- Check auth: Call
check_auth_statusto verify the service account is configured - List accounts: Call
list_accountsto see your GA4 accounts - Setup a property: Use
create_property→create_web_data_stream→get_tracking_snippet - Query data: Use
run_reportwith dimensions and metrics
Tools (23)
Auth
| Tool | Description |
|---|---|
check_auth_status |
Verify service account is configured and valid |
Admin API
| Tool | Description |
|---|---|
list_accounts |
List accessible GA4 accounts |
list_properties |
List properties for an account |
create_property |
Create a GA4 property |
get_property |
Get property details |
delete_property |
Soft-delete a property |
list_data_streams |
List data streams |
create_web_data_stream |
Create web stream → get Measurement ID |
delete_data_stream |
Delete a data stream |
get_tracking_snippet |
Generate gtag.js HTML snippet |
list_key_events |
List conversion events |
create_key_event |
Mark event as conversion |
delete_key_event |
Remove conversion |
list_custom_dimensions |
List custom dimensions |
create_custom_dimension |
Create custom dimension |
list_custom_metrics |
List custom metrics |
create_custom_metric |
Create custom metric |
list_google_ads_links |
List Google Ads links |
create_google_ads_link |
Link Google Ads account |
get_measurement_protocol_secret |
Get/create Measurement Protocol secret |
Data API
| Tool | Description |
|---|---|
run_report |
Run report with dimensions, metrics, dates, filters |
run_realtime_report |
Real-time active users and events |
get_metadata |
List available dimensions and metrics |
Environment Variables
| Variable | Default | Description |
|---|---|---|
GA_CREDENTIALS |
— | Service Account JSON key content (copy-paste the entire JSON) |
GA_CREDENTIALS_PATH |
~/.google-analytics-4-mcp/credentials.json |
Alternative: path to the JSON key file |
GA_CREDENTIALS takes priority over GA_CREDENTIALS_PATH.
Example Usage
You: "Set up Google Analytics for my website https://mcpadvisor.stucchi.consulting"
Assistant calls: list_accounts → create_property → create_web_data_stream → get_tracking_snippet
Result: Property created with Measurement ID G-XXXXXXXXXX and ready-to-paste HTML snippet.
You: "Show me the top 10 countries by sessions in the last 7 days"
Assistant calls: run_report(property_id="...", dimensions=["country"], metrics=["sessions"], start_date="7daysAgo", limit=10)
License
MIT
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。