imessage-kit-mcp
Enables reading, sending, and managing iMessage conversations on macOS through MCP.
README
iMessage Kit MCP Server
An MCP server for iMessage — read, send, and automate conversations via iMessage Kit.
Features
- Send Messages: Send iMessages to phone numbers or email addresses
- Read Messages: Retrieve message history from conversations
- List Conversations: Get all active iMessage conversations
- Conversation Details: Get detailed information about specific conversations
- Dual Transport: Supports both STDIO (local) and HTTP (production) transports
Tools
imsg_send_message
Sends an iMessage to a recipient.
Inputs:
recipient(string, required): Phone number (e.g., '+1234567890') or email addresstext(string, required): Message text to sendchatId(string, optional): Chat ID if continuing an existing conversation
imsg_read_messages
Reads messages from an iMessage conversation.
Inputs:
chatId(string, optional): Chat/conversation IDrecipient(string, optional): Recipient phone number or emaillimit(number, optional): Maximum number of messages (default: 50)offset(number, optional): Number of messages to skip (default: 0)
imsg_get_conversations
Gets a list of all iMessage conversations.
Inputs:
limit(number, optional): Maximum number of conversations (default: 50)offset(number, optional): Number of conversations to skip (default: 0)
imsg_get_conversation_details
Gets detailed information about a specific conversation.
Inputs:
chatId(string, required): Chat/conversation ID
Installation
Prerequisites
- Bun — required runtime
- macOS (required for iMessage functionality)
Install Dependencies
bun install
Configuration
Usage with Claude Desktop
Add this to your claude_desktop_config.json:
STDIO Transport (Local)
{
"mcpServers": {
"imessage-kit": {
"command": "bun",
"args": ["run", "/path/to/imessage-kit-mcp/src/index.ts"]
}
}
}
HTTP Transport
{
"mcpServers": {
"imessage-kit": {
"command": "bun",
"args": ["run", "/path/to/imessage-kit-mcp/src/index.ts", "--port", "3002"],
"url": "http://localhost:3002/mcp"
}
}
}
Usage with VS Code
Add the following JSON block to your User Settings (JSON) file in VS Code. You can do this by pressing Ctrl + Shift + P and typing Preferences: Open User Settings (JSON).
{
"mcp": {
"inputs": [],
"servers": {
"imessage-kit": {
"command": "bun",
"args": ["run", "/absolute/path/to/imessage-kit-mcp/src/index.ts"]
}
}
}
}
Development
Running the Server
# STDIO transport (default)
bun run src/index.ts
# HTTP transport
bun run src/index.ts --port 3002
Testing with MCP Inspector
bunx @modelcontextprotocol/inspector src/index.ts
Project Structure
imessage-kit-mcp/
├── src/
│ ├── index.ts # Main entry point
│ ├── server.ts # MCP server implementation
│ ├── client.ts # iMessage Kit SDK client wrapper
│ ├── config.ts # Configuration management
│ ├── cli.ts # CLI argument parsing
│ ├── types.ts # TypeScript type definitions
│ ├── tools/
│ │ ├── index.ts # Tool exports
│ │ └── imessage.ts # iMessage tool definitions
│ └── transport/
│ ├── index.ts # Transport exports
│ ├── stdio.ts # STDIO transport
│ └── http.ts # HTTP transport
├── package.json
├── tsconfig.json
└── README.md
Notes
- This server requires macOS to access iMessage functionality
- Requires Bun (uses
bun:sqlitefor native SQLite access) - Uses iMessage Kit for database access
- Make sure you have proper permissions to access Messages on macOS
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。