ntfy-mcp

ntfy-mcp

MCP server for sending notifications to ntfy.sh or self-hosted ntfy instances.

Category
访问服务器

README

ntfy-mcp

Standalone MCP server for sending notifications to ntfy.sh or any self-hosted ntfy instance.

This server is reusable across MCP clients. It exposes tools for sending notifications, while keeping the ntfy URL and topic locked at server startup so agents cannot override them per call.

Features

  • stdio MCP server for broad client compatibility
  • locked ntfy base URL and topic
  • startup config via flags or environment variables
  • bearer token or basic auth support
  • ntfy_publish tool for general notifications
  • ntfy_ping tool for simple "finished working" notifications

Install

Local development

npm install
npm run build

Published CLI

After publishing, clients can run the server directly with npx:

npx -y ntfy-mcp --ntfy-url https://ntfy.sh --ntfy-topic my-topic

Smoke test

You can run a local MCP-level smoke test after building:

node scripts/smoke-test.mjs dist/index.js \
  --ntfy-url https://ntfy.sh \
  --ntfy-topic my-test-topic

Configuration

Flags take priority over environment variables.

Flags

node dist/index.js \
  --ntfy-url https://ntfy.sh \
  --ntfy-topic my-topic \
  --ntfy-token tk_your_token

Environment variables

export NTFY_URL="https://ntfy.sh"
export NTFY_TOPIC="my-topic"
export NTFY_TOKEN="tk_your_token"
node dist/index.js

Supported variables:

  • NTFY_URL
  • NTFY_TOPIC
  • NTFY_TOKEN
  • NTFY_USERNAME
  • NTFY_PASSWORD

If both token auth and basic auth are configured, token auth wins.

Tool reference

ntfy_publish

Sends a notification to the configured topic.

Input:

{
  "message": "Deployment complete",
  "title": "Deploy",
  "priority": "high",
  "tags": ["rocket", "white_check_mark"],
  "markdown": false,
  "click": "https://example.com/deploys/123",
  "icon": "https://example.com/icon.png",
  "delay": "10m"
}

ntfy_ping

Convenience tool for simple pings.

Input:

{
  "message": "Agent finished working.",
  "title": "Agent finished",
  "priority": "default",
  "tags": ["robot_face", "white_check_mark"]
}

npm publish

This package is set up to publish as a public npm CLI.

npm login
npm publish

Quick checks before publishing:

npm run build
npm pack --dry-run

skills.sh skill

This repository also contains a reusable skill in the skills/ directory:

  • skills/ntfy-after-task/SKILL.md

After pushing the repo, it can be installed through skills.sh with:

npx skills add Edqe14/ntfy-mcp@ntfy-after-task

OpenCode

Register the server in opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "ntfy": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "ntfy-mcp",
        "--ntfy-url",
        "https://ntfy.sh",
        "--ntfy-topic",
        "my-topic"
      ]
    }
  }
}

This only makes the tool available. To send a ping after the agent finishes working, OpenCode still needs a plugin, hook, or workflow that calls ntfy_ping when the session becomes idle or reaches your chosen completion event.

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "ntfy": {
      "command": "npx",
      "args": [
        "-y",
        "ntfy-mcp",
        "--ntfy-url",
        "https://ntfy.sh",
        "--ntfy-topic",
        "my-topic"
      ]
    }
  }
}

Cursor

Add to .cursor/mcp.json or ~/.cursor/mcp.json:

{
  "mcpServers": {
    "ntfy": {
      "command": "npx",
      "args": [
        "-y",
        "ntfy-mcp",
        "--ntfy-url",
        "https://ntfy.sh",
        "--ntfy-topic",
        "my-topic"
      ]
    }
  }
}

Notes

  • The server publishes with HTTP POST to https://<ntfy-host>/<topic>.
  • ntfy URL and topic are intentionally locked at startup.
  • Per-call overrides for URL or topic are not supported.
  • Use console.error for logs because stdout is reserved for MCP protocol traffic.

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