clarity-mcp-multi
Enables managing multiple Microsoft Clarity projects by routing requests per client name using a token map, proxying the official Clarity MCP server with tools to list and switch clients.
README
clarity-mcp-multi
An MCP proxy that adds per-client token routing on top of Microsoft's official Clarity MCP server.
Why
The official server authenticates with a single Clarity Data Export token, and a token maps to exactly one project. If you manage several Clarity projects you end up editing the config and restarting the client every time you switch project.
This proxy keeps a name → token map and resolves the token by name at runtime. It does not reimplement any Clarity functionality: it launches the official server as a child process and forwards calls to it, so it inherits the full tool set and tracks upstream changes automatically.
How it works
The package acts as a server towards its MCP client (Claude, Cursor, …) and as a client towards the official server at the same time. It registers two tools of its own and re-exports whatever the official server exposes. A call that isn't one of its own tools is forwarded to the child process and the response is returned unchanged.
Selecting a client with clarity_usa_cliente restarts the child with that client's token. The token is read from the registry on the server side and is never sent through the conversation.
The child is launched as @microsoft/clarity-mcp-server@latest, so upstream releases are picked up without changes here.
Configuration
{
"mcpServers": {
"clarity": {
"command": "npx",
"args": ["-y", "clarity-mcp-multi@latest"],
"env": {
"CLARITY_TOKENS": "{\"acme\":\"<token>\",\"globex\":\"<token>\"}"
}
}
}
}
CLARITY_TOKENS is a JSON object mapping a client name to its Clarity Data Export token (Clarity → Settings → Data export). As an alternative, set CLARITY_TOKENS_FILE to a path holding the same JSON. CLARITY_DEFAULT_CLIENT optionally sets the client selected at startup.
Keep tokens out of source control: pass them through the environment or a file the host loads, not in a committed config.
Tools
| Tool | Description |
|---|---|
clarity_clienti |
Lists the configured client names and the active one. No token is returned. |
clarity_usa_cliente |
Sets the active client by name; subsequent calls use its token. |
Names are matched case-insensitively. Every other tool is proxied to the official server under the active client's token.
Requirements
Node.js 18 or later. One Clarity Data Export token per project.
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。