GrowthBook MCP Server

Name: GrowthBook MCP Server
Author: growthbook

@growthbook

代码仓库主页

关于 GrowthBook MCP Server

Official GrowthBook MCP server for creating flags, getting experiments, and more.

#growthbook #feature-flags #experimentation

基本信息

分类

其他

许可证

MIT

运行时

node

传输方式

stdio

发布者

growthbook

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "growthbook": {
      "command": "npx",
      "args": [
        "-y",
        "@growthbook/mcp@latest"
      ],
      "env": {
        "GB_API_KEY": "YOUR_API_KEY",
        "GB_API_URL": "YOUR_API_URL",
        "GB_APP_ORIGIN": "YOUR_APP_ORIGIN",
        "GB_USER": "YOUR_NAME"
      }
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is GrowthBook MCP Server?

GrowthBook MCP Server lets you interact with GrowthBook directly from an LLM client. It is designed for users who want to view experiment details, add feature flags, and perform other GrowthBook actions using natural language commands.

How to use GrowthBook MCP Server?

Add the MCP server to your AI tool of choice and configure it using environment variables. Required variables are GB_API_KEY (a GrowthBook API key or PAT) and GB_EMAIL (your email). Optional variables include GB_API_URL, GB_APP_ORIGIN, and GB_HTTP_HEADER_* for custom headers. See the official GrowthBook docs for a complete setup guide.

Core features of GrowthBook MCP Server

View experiment details from GrowthBook
Create and manage feature flags
Create new experiments
Integrate with any LLM client
Configure via environment variables
Support for custom HTTP headers

Use cases of GrowthBook MCP Server

Query experiment results and feature flag status through chat
Create a new feature flag using natural language commands
Manage growthbook experiments from within an LLM interface
Configure enterprise proxy or tenant headers for GrowthBook API

FAQ from GrowthBook MCP Server

What runtime or dependencies does GrowthBook MCP Server require?

The server is an MCP server and must be added to an LLM client that supports the Model Context Protocol. It requires no specific runtime beyond the client's environment.

How is authentication handled?

Authentication uses a GrowthBook API key or Personal Access Token (PAT) set via the GB_API_KEY environment variable. When using a PAT, server capabilities are limited by the token's permissions.

Where does data live?

All data interactions go through the GrowthBook API. Feature flags, experiments, and user data remain within GrowthBook's infrastructure.

What are the known limits or restrictions?

Capabilities are limited by the permissions of the API key or PAT used. For example, if the token cannot create experiments in the GrowthBook app, it also cannot create them via the MCP server.

What transport or protocol does GrowthBook MCP Server use?

The server uses the Model Context Protocol (MCP) to communicate with LLM clients. It connects to GrowthBook via its REST API, with the API URL configurable via the GB_API_URL environment variable.