posthog-mcp

posthog-mcp

Enables AI agents to query PostHog analytics data directly via tool calls, including insights, events, feature flags, trends, and persons.

Category
访问服务器

README

posthog-mcp

An MCP (Model Context Protocol) server that wraps the PostHog REST API, letting AI agents query your analytics data directly via tool calls — no browser tab required.

Works with Claude Desktop, Cursor, and any other MCP-compatible client.

What it does

Exposes five tools that cover the most common PostHog workflows:

Tool What it does
get_insights Fetch a saved insight by ID (funnels, retention, trends)
list_events Stream recent events, filterable by name, date range, or distinct ID
get_feature_flags List all flags with rollout status and targeting rules
query_trends Run an ad-hoc trends query and get a time series back
get_persons Look up user profiles by distinct ID or search term

Prerequisites

  • Node.js 18+
  • A PostHog account with a personal API key (Settings → Personal API Keys)
  • Your project ID (visible in Project Settings or in the URL: /project/<id>/)

Installation

git clone https://github.com/johnnikolo/posthog-mcp.git
cd posthog-mcp
npm install
cp .env.example .env   # then fill in your API key and project ID
npm run build

Configuration

The server is configured via environment variables. When run locally, it automatically loads a .env file in the project root (via dotenv); when configured through an MCP client, use the client's env block instead (see below).

Variable Required Default Description
POSTHOG_API_KEY Yes — Your PostHog personal API key
POSTHOG_PROJECT_ID Yes — The numeric project ID to query
POSTHOG_HOST No https://app.posthog.com Override for self-hosted PostHog instances

Wiring it up in Claude Desktop

Add this to your claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json, Windows: %APPDATA%\Claude\claude_desktop_config.json):

{
  "mcpServers": {
    "posthog": {
      "command": "node",
      "args": ["/absolute/path/to/posthog-mcp/dist/index.js"],
      "env": {
        "POSTHOG_API_KEY": "phx_your_key_here",
        "POSTHOG_PROJECT_ID": "12345"
      }
    }
  }
}

Restart Claude Desktop and the PostHog tools will appear automatically.

Tools reference

get_insights

Retrieve a saved PostHog insight by its numeric ID.

What does insight 42 show?
→ calls get_insights({ insight_id: 42 })

Parameters

  • insight_id (number, required) — the insight's ID
  • refresh (boolean, default false) — force a fresh calculation

list_events

Fetch recent events with optional filters.

Show me the last 50 user_signed_up events from the past week
→ calls list_events({ event: "user_signed_up", after: "2024-01-01T00:00:00Z", limit: 50 })

Parameters

  • event (string) — filter to a specific event name
  • after / before (ISO 8601 string) — date range
  • limit (number, default 100, max 500)
  • distinct_id (string) — filter to a single user

get_feature_flags

List all feature flags with their rollout configuration.

Which feature flags are currently active?
→ calls get_feature_flags({ active_only: true })

Parameters

  • active_only (boolean, default false) — return only enabled flags

query_trends

Run a trends query and get back a time series.

How many $pageview and user_signed_up events happened per day over the last 30 days?
→ calls query_trends({ events: ["$pageview", "user_signed_up"], date_from: "-30d", interval: "day" })

Parameters

  • events (string[], required) — event names to include
  • date_from (string, default "-7d") — start of range (relative or ISO 8601)
  • date_to (string) — end of range (defaults to now)
  • interval ("hour" | "day" | "week" | "month", default "day")
  • breakdown (string) — property to break results down by (e.g. "$browser")

get_persons

Look up person profiles.

Find the user with distinct ID "user_abc123"
→ calls get_persons({ distinct_id: "user_abc123" })

Parameters

  • distinct_id (string) — look up a specific person
  • search (string) — search by email, name, or property
  • limit (number, default 20, max 100)

Development

# run directly without building
npm run dev

# type-check only
npm run typecheck

License

MIT

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选