MCP Connector

Send Notifications Across Multiple Platforms

Send notifications from AI agents to WeCom, DingTalk, Telegram, Bark, Lark/Feishu, Ntfy, PushPlus, or Home Assistant via one MCP server.

Works with wechatdingtalktelegramhome assistant

Maintainer of this project? Claim this page to edit the listing.


91
Spark score
out of 100
Updated 4 months ago
Version 0.1.11
Models
universal

Add to Favorites

Why it matters

Integrate your applications and systems with a unified messaging server to send notifications across WeChat Work, DingTalk, Telegram, and more. Automate alerts and updates to streamline your DevOps and communication workflows.

Outcomes

What it gets done

01

Configure and deploy a notification server via Docker, uvx, or Smithery.

02

Send text, image, video, audio, and file messages to various platforms.

03

Manage API keys and secrets for secure communication.

04

Leverage Markdown formatting for enhanced message content.

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/vb-mcp-notify | bash

Capabilities

Tools your agent gets

wework_send_text

Send text or Markdown messages via WeChat Work group bot

wework_send_image

Send images via WeChat Work group bot

wework_send_news

Send news messages via WeChat Work group bot

wework_app_send_text

Send text or Markdown messages via WeChat Work app

wework_app_send_image

Send images via WeChat Work app

wework_app_send_video

Send videos via WeChat Work app

wework_app_send_voice

Send voice messages via WeChat Work app

wework_app_send_file

Send files via WeChat Work app

+7 tools

Overview

mcp-notify MCP Server

mcp-notify is an MCP server that sends notifications from an AI agent to WeCom, DingTalk, Feishu/Lark, Telegram, Bark, Ntfy, PushPlus, and Home Assistant. It supports text, image, video, voice, and file messages depending on the platform. Use it when an AI agent needs to push an alert or message into an existing chat or notification channel. Not applicable for two-way conversation or platforms outside the supported list.

What it does

mcp-notify is an MCP server dedicated to pushing notifications from an AI assistant to a wide range of messaging and IoT platforms: WeCom (Enterprise WeChat) group bots and app accounts, DingTalk, Feishu/Lark, Telegram, Bark, Ntfy, PushPlus, and Home Assistant mobile notifications.

When to use - and when NOT to

Use it when you want an AI agent to alert you or a team through an existing chat or notification channel - for example, pushing a task-completion message to a WeCom or DingTalk group, sending a Telegram message with a photo or file, or triggering a Home Assistant mobile push. Credentials (bot keys, tokens, chat IDs) can be set as environment variables as defaults, or specified directly in the prompt per call. Do not use it if you need bidirectional chat (this is send-only notification, not conversation) or if none of the supported platforms match your notification channel.

Capabilities

WeCom group bot: wework_send_text (text/Markdown), wework_send_image, wework_send_news (article links).
WeCom app account: wework_app_send_text, wework_app_send_image, wework_app_send_video, wework_app_send_voice, wework_app_send_file, wework_app_send_news.
Telegram: tg_send_message, tg_send_photo, tg_send_video, tg_send_audio, tg_send_file.
Other channels: ding_send_text (DingTalk group bot), lark_send_text (Feishu/Lark group bot), bark_send_notify, ntfy_send_notify, pushplus_send_msg, ha_send_mobile (Home Assistant), and text_to_sound (convert text to an MP3 audio link).

Each platform is configured through its own set of environment variables, e.g. WEWORK_BOT_KEY for the WeCom group bot, WEWORK_APP_CORPID/WEWORK_APP_SECRET/WEWORK_APP_AGENTID for the WeCom app, DINGTALK_BOT_KEY, FEISHU_BOT_KEY/LARK_BOT_KEY, TELEGRAM_BOT_TOKEN/TELEGRAM_DEFAULT_CHAT, HASS_BASE_URL/HASS_ACCESS_TOKEN/HASS_MOBILE_KEY, BARK_DEVICE_KEY, NTFY_DEFAULT_TOPIC, and PUSH_PLUS_TOKEN - most also support a custom _BASE_URL for self-hosted or proxied API endpoints.

How to install

Via uvx:

{
  "mcpServers": {
    "mcp-notify": {
      "command": "uvx",
      "args": ["mcp-notify"],
      "env": { "WEWORK_BOT_KEY": "your-wework-bot-key" }
    }
  }
}

Or via Smithery's hosted endpoint (OAuth or Smithery key required), or self-hosted with Docker:

mkdir /opt/mcp-notify && cd /opt/mcp-notify
wget https://raw.githubusercontent.com/aahl/mcp-notify/refs/heads/main/docker-compose.yml
docker-compose up -d

For Claude Code: claude mcp add notify -- uvx mcp-notify. For OpenAI Codex: codex mcp add notify -- uvx mcp-notify. One-click install links are also provided for Cursor, VS Code, and Cherry Studio.

Who it's for

Developers and automation builders who want an AI agent to push alerts, reports, or media into their existing WeCom, DingTalk, Feishu/Lark, Telegram, Bark, Ntfy, PushPlus, or Home Assistant channels.

Source README

💬 Notify MCP Server

简体中文 | English

提供消息推送的 MCP (Model Context Protocol) 服务器,支持企业微信、钉钉、Telegram、Bark、Lark、飞书、Home Assistant

安装

方式1: uvx

{
  "mcpServers": {
    "mcp-notify": {
      "command": "uvx",
      "args": ["mcp-notify"],
      "env": {
        "WEWORK_BOT_KEY": "your-wework-bot-key"
      }
    }
  }
}

方式2: Smithery

需要通过OAuth授权或Smithery key

{
  "mcpServers": {
    "mcp-aktools": {
      "url": "https://server.smithery.ai/@aahl/mcp-notify/mcp" # 流式传输HTTP
    }
  }
}

方式3: Docker

mkdir /opt/mcp-notify
cd /opt/mcp-notify
wget https://raw.githubusercontent.com/aahl/mcp-notify/refs/heads/main/docker-compose.yml
docker-compose up -d
{
  "mcpServers": {
    "mcp-notify": {
      "url": "http://0.0.0.0:8809/mcp" # 流式传输HTTP
    }
  }
}

快速开始

  • 在线体验: fastmcp.cloud
  • 在线体验: smithery.ai
  • 添加到 Cursor Install MCP Server
  • 添加到 VS Code Install MCP Server
  • 添加到 Cherry Studio Install MCP Server
  • 添加到 Claude Code, 执行命令: claude mcp add notify -- uvx mcp-notify
  • 添加到 OpenAI CodeX, 执行命令: codex mcp add notify -- uvx mcp-notify

⚙️ 环境变量

企业微信群机器人
  • WEWORK_BOT_KEY: 企业微信群机器人默认key,也可以在提示词指定
企业微信应用号
  • WEWORK_APP_CORPID: 企业微信所属的企业ID
  • WEWORK_APP_SECRET: 企业微信应用的凭证密钥
  • WEWORK_APP_AGENTID: 企业微信应用的ID,默认: 1000002
  • WEWORK_APP_TOUSER: 企业微信默认接收人ID,也可以在提示词指定,默认: @all
  • WEWORK_BASE_URL: 企业微信API反代理地址,用于可信IP,默认: https://qyapi.weixin.qq.com
钉钉群机器人
  • DINGTALK_BOT_KEY: 钉钉群机器人access_token
  • DINGTALK_BASE_URL: 钉钉API地址,默认: https://oapi.dingtalk.com
飞书/Lark群机器人
  • FEISHU_BOT_KEY: 飞书群机器人key,也可以在提示词指定
  • FEISHU_BASE_URL: 飞书API地址,默认: https://open.feishu.cn
  • LARK_BOT_KEY: Lark群机器人key,也可以在提示词指定
  • LARK_BASE_URL: Lark API地址,默认: https://open.larksuite.com
Telegram
  • TELEGRAM_DEFAULT_CHAT: Telegram 默认会话ID,也可以在提示词指定
  • TELEGRAM_BOT_TOKEN: Telegram 机器人令牌
  • TELEGRAM_BASE_URL: Telegram API反代理地址,默认: https://api.telegram.org
Home Assistant
  • HASS_BASE_URL: Home Assistant 地址,默认: http://homeassistant.local:8123
  • HASS_ACCESS_TOKEN: Home Assistant 长效令牌
  • HASS_MOBILE_KEY: Home Assistant 移动设备key (如: mobile_app_your_iphone),也可在提示词指定
其他
  • BARK_DEVICE_KEY: 默认Bark设备key,也可以在提示词指定
  • BARK_BASE_URL: Bark API地址,默认: https://api.day.app
  • NTFY_DEFAULT_TOPIC: 默认Ntfy订阅主题,也可以在提示词指定
  • NTFY_BASE_URL: Ntfy API地址,默认: https://ntfy.sh
  • PUSH_PLUS_TOKEN: 默认PushPlus令牌,也可以在提示词指定
  • PUSH_PLUS_BASE_URL: PushPlus API地址,默认: http://www.pushplus.plus

🛠️ 可用工具

企业微信群机器人
  • wework_send_text - 发送文本或Markdown消息
  • wework_send_image - 发送图片消息
  • wework_send_news - 发送图文链接消息
企业微信应用号
  • wework_app_send_text - 发送文本或Markdown消息
  • wework_app_send_image - 发送图片消息
  • wework_app_send_video - 发送视频消息
  • wework_app_send_voice - 发送语音消息
  • wework_app_send_file - 发送文件消息
  • wework_app_send_news - 发送图文链接消息
Telegram Bot
  • tg_send_message - 发送文本或Markdown消息
  • tg_send_photo - 发送图片消息
  • tg_send_video - 发送视频消息
  • tg_send_audio - 发送音频消息
  • tg_send_file - 发送文件消息
其他工具
  • ding_send_text - 通过钉钉群机器人发送文本或Markdown消息
  • lark_send_text - 通过飞书/Lark群机器人发送文本或Markdown消息
  • bark_send_notify - 通过Bark发送通知
  • ntfy_send_notify - 通过Ntfy发送通知
  • pushplus_send_msg - 通过PushPlus发送消息
  • ha_send_mobile - 通过Home Assistant发送通知
  • text_to_sound - 将一段文本转成mp3音频链接

🔗 相关连接


FAQ

Common questions

Discussion

Questions & comments · 0

Sign In Sign in to leave a comment.