HitKeep
Read-only MCP server for aggregate HitKeep analytics and official documentation, including traffic, events, goals, funnels, ecommerce, Search Console, and AI visibility reporting.
README
HitKeep
Privacy-first analytics for humans and AI agents, self-hosted or in EU/US cloud.
HitKeep is open source web analytics for teams that want useful product reporting without running PostgreSQL, Redis, ClickHouse, or a separate queue.
- Single Go binary with embedded DuckDB and NSQ
- Cookie-less tracker by default, with Do Not Track support
- Traffic, events, goals, funnels, ecommerce, UTM, and email reports
- Google Search Console aggregate import for query, page, country, and device reporting
- AI visibility analytics for crawler fetches and AI-referred visits
- Scoped API clients and a read-only MCP analytics server for approved agents
- Self-hosted or managed cloud with EU/US region choice
AI Performance · Website · Cloud · Live Demo · Docs · API · Releases

Why HitKeep
HitKeep is for teams that need clear web analytics, conversion reporting, and AI-era search visibility in one small operational footprint.
- Low-ops self-hosting: one binary, one data directory, embedded DuckDB and NSQ
- Useful reports: top pages, landing and exit pages, events, goals, funnels, ecommerce, UTM attribution, and Search Console aggregates
- Privacy defaults: focused data collection, cookie-less tracking, and DNT handling
- AI visibility: server-side crawler fetch analytics, AI-referred visits, and correlation reports
- Team controls: passkeys, TOTP, site/team permissions, share links, audit logs, API clients, and read-only MCP access
- Deployment choice: run it yourself or use managed cloud in the EU or US
Quick Start
Binary
Download the latest release for your system:
wget https://github.com/PascaleBeier/hitkeep/releases/latest/download/hitkeep-linux-arm64
chmod +x hitkeep-linux-arm64
export HITKEEP_JWT_SECRET="replace-this-with-a-long-random-string"
./hitkeep-linux-arm64 -public-url="http://localhost:8080"
Open http://localhost:8080 and create your first account.
Docker
services:
hitkeep:
image: pascalebeier/hitkeep:latest
restart: unless-stopped
ports:
- "8080:8080"
volumes:
- hitkeep_data:/var/lib/hitkeep/data
environment:
HITKEEP_JWT_SECRET: replace-this-with-a-long-random-string
command:
- "-public-url=http://localhost:8080"
volumes:
hitkeep_data: {}
docker compose up -d
For production setup, reverse proxies, SMTP, systemd, Kubernetes, S3 archiving, and every configuration flag, use the docs instead of this README:
Local Development
Contributors can run the full hot-reload development stack with Docker:
make dev-docker-seed
This starts the Go backend, Angular dashboard, Mailpit, and seeded demo data.
Open http://localhost:4200 and sign in with demo@example.com / demo1234.
See CONTRIBUTING.md for Docker-only and native workflows.
Track Your Site
Once your instance is running and a site is created, add:
<script async src="https://your-hitkeep-instance.com/hk.js"></script>
Custom event example:
<script>
window.hk = window.hk || {};
window.hk.event?.("signup", { plan: "pro", source: "landing-page" });
</script>
Tracker options, ecommerce events, custom events, and advanced tracking examples live here:
- Tracking docs
- Custom events
- Ecommerce analytics
- Google Search Console integration
- Read-only MCP server for web analytics
- WordPress integration
- AI visibility analytics
- CloudFront AI crawler tracking
- AI chatbot analytics
- REST API reference
- Compliance overview
Product Tour
<details> <summary>See five product screenshots</summary>
Dashboard

Ecommerce

Search Console

AI Visibility

MCP Access

</details>
Documentation
The maintained reference lives on hitkeep.com.
- Getting started
- Installation
- Configuration
- REST API reference
- Compliance
- Comparison pages
- Agent Skills
Cloud
If you want the same product without running it yourself, start here:
Development
Prerequisites:
- Go 1.26+
- Node.js 24+
- Make
- A working C toolchain for DuckDB builds
Build from source:
git clone https://github.com/pascalebeier/hitkeep.git
cd hitkeep
make build
./hitkeep
For day-to-day development:
make dev
This starts the Go backend with live reload and the Angular dashboard on http://localhost:4200.
For a seeded local workspace with demo data:
make dev-seed
Contributor docs and local development guides:
License
Distributed under the MIT License. See LICENSE.
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。