BetterTouchTool MCP Server

BetterTouchTool MCP Server

Enables AI assistants to control and interact with BetterTouchTool on macOS, allowing users to query configurations, execute triggers and actions, manage variables and widgets, control presets, and access clipboard contents through natural language.

Category
访问服务器

README

BetterTouchTool MCP Server

A Model Context Protocol (MCP) server that enables AI assistants like Claude to control and interact with BetterTouchTool on macOS.

What is this?

This MCP server acts as a bridge between Claude (or other MCP-compatible AI assistants) and BetterTouchTool, allowing you to:

  • Query your BTT configuration - Understand what triggers, shortcuts, and automations you have set up
  • Execute triggers and actions - Run named triggers, keyboard shortcuts, or any BTT action
  • Manage variables - Get and set BTT variables for dynamic workflows
  • Update widgets - Modify Touch Bar, Menu Bar, or Stream Deck widgets in real-time
  • Control presets - Export, import, and manage BTT presets
  • Access clipboard - Read and write clipboard contents programmatically

Prerequisites

  • macOS (BetterTouchTool is macOS-only)
  • BetterTouchTool installed with a valid license
  • Python 3.12+
  • uv (recommended) or pip for package management
  • Claude Desktop or another MCP-compatible client

Installation

1. Enable BTT Webserver

First, enable the BetterTouchTool webserver:

  1. Open BetterTouchTool
  2. Go to PreferencesAdvancedWebserver
  3. Check Enable Webserver
  4. Note the port (default: 12345)
  5. Optionally set a Shared Secret for authentication

2. Install the MCP Server

Option A: Install with uv (Recommended)

# Clone the repository
git clone https://github.com/your-username/bettertouchtool-mcp-server.git
cd bettertouchtool-mcp-server

# Install dependencies with uv
uv sync

Option B: Install with pip

# Clone the repository
git clone https://github.com/your-username/bettertouchtool-mcp-server.git
cd bettertouchtool-mcp-server

# Create virtual environment and install
python -m venv .venv
source .venv/bin/activate
pip install -e .

3. Configure Claude Desktop

Add the server to your Claude Desktop configuration file:

Location: ~/Library/Application Support/Claude/claude_desktop_config.json

Using uv:

{
  "mcpServers": {
    "bettertouchtool": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/bettertouchtool-mcp-server",
        "run",
        "btt_mcp.py"
      ]
    }
  }
}

Using Python directly:

{
  "mcpServers": {
    "bettertouchtool": {
      "command": "/path/to/bettertouchtool-mcp-server/.venv/bin/python",
      "args": ["/path/to/bettertouchtool-mcp-server/btt_mcp.py"]
    }
  }
}

4. Restart Claude Desktop

After saving the configuration, restart Claude Desktop to load the MCP server.

Available Tools

The server provides 20 tools organized into categories:

Configuration Reading

Tool Description
btt_get_triggers List triggers with optional filtering by type, app, or parent
btt_get_trigger Get detailed info about a specific trigger by UUID
btt_list_named_triggers List all named triggers (quick actions)
btt_get_preset_details Get info about presets and their status

Trigger Management

Tool Description
btt_trigger_named Execute a named trigger by name
btt_trigger_action Execute any BTT action via JSON definition
btt_execute_trigger Execute all actions assigned to a trigger UUID
btt_add_trigger Create a new trigger from JSON definition
btt_update_trigger Modify an existing trigger
btt_delete_trigger Remove a trigger (⚠️ destructive)

Variable Management

Tool Description
btt_get_variable Read string or number variables
btt_set_variable Set persistent or runtime variables

Widget Control

Tool Description
btt_update_widget Update widget text, icon, or colors
btt_refresh_widget Force a script widget to refresh

Clipboard Operations

Tool Description
btt_get_clipboard Read clipboard in various formats
btt_set_clipboard Write to clipboard

Preset Management

Tool Description
btt_export_preset Export preset to a file
btt_import_preset Import preset from a file

System Utilities

Tool Description
btt_display_notification Show a macOS notification
btt_reveal_in_ui Open BTT and navigate to an element

Usage Examples

Exploring Your Configuration

Ask Claude:

"What triggers do I have set up in BetterTouchTool?"

"Show me all my named triggers"

"What keyboard shortcuts do I have configured?"

Running Automations

"Run my 'Toggle Dark Mode' trigger in BTT"

"Execute the BTT trigger that opens my development environment"

Working with Variables

"What's the current value of my 'workMode' variable in BTT?"

"Set my BTT variable 'currentProject' to 'MCP Server Development'"

Updating Widgets

"Update my Touch Bar clock widget to show 'Meeting in 5 min' with a red background"

"Refresh my system stats widget"

Managing Configuration

"Export my current BTT preset to ~/Desktop/my-preset.bttpreset"

"Show me the details of trigger with UUID abc-123-def"

Complex Workflows

"I want to create a new named trigger that opens Safari and navigates to my project management tool"

"Help me understand what my 'Morning Routine' trigger does and suggest improvements"

Connection Options

Default HTTP Connection

By default, tools connect via HTTP to 127.0.0.1:12345. No additional configuration needed if BTT webserver is enabled with default settings.

With Shared Secret

If you've configured a shared secret in BTT, include it in your requests:

Use my BTT shared secret "my_secret_key" when connecting

Or Claude will ask if authentication fails.

Using CLI Mode

For faster performance, you can use the bttcli command-line tool instead of HTTP:

Use CLI mode for BTT commands

This uses the Unix socket at /tmp/com.hegenberg.BetterTouchTool.sock for lower latency.

Trigger Types

When filtering triggers, you can use these types:

  • named_trigger - Named triggers (quick actions)
  • keyboard_shortcut - Keyboard shortcuts
  • touchbar - Touch Bar items
  • trackpad - Trackpad gestures
  • magic_mouse - Magic Mouse gestures
  • drawings - Drawing gestures
  • siri_remote - Siri Remote
  • normal_mouse - Regular mouse buttons
  • stream_deck - Stream Deck buttons
  • midi - MIDI triggers
  • menubar - Menu bar items

Troubleshooting

"Connection refused" errors

  1. Ensure BTT is running
  2. Verify webserver is enabled in BTT preferences
  3. Check the port matches (default: 12345)

"403 Forbidden" errors

You have a shared secret configured in BTT. Either:

  • Provide the secret in your request
  • Remove the shared secret from BTT preferences

"bttcli not found" (CLI mode)

The CLI tool should be at:

  • /Applications/BetterTouchTool.app/Contents/SharedSupport/bin/bttcli
  • or ~/Applications/BetterTouchTool.app/Contents/SharedSupport/bin/bttcli

Tools not appearing in Claude

  1. Check Claude Desktop config syntax is valid JSON
  2. Verify the path to the MCP server is correct
  3. Restart Claude Desktop
  4. Check Console.app for MCP-related errors

Development

Running locally for testing

# With uv
uv run btt_mcp.py

# Or directly
python btt_mcp.py

Testing with MCP Inspector

npx @modelcontextprotocol/inspector uv --directory /path/to/project run btt_mcp.py

API Reference

This server implements the BTT Webserver API. Key endpoints used:

  • GET /get_triggers - Retrieve trigger configurations
  • GET /trigger_named - Execute named triggers
  • GET /trigger_action - Execute arbitrary actions
  • GET /get_string_variable / set_string_variable - Variable management
  • GET /update_touch_bar_widget - Widget updates
  • GET /get_clipboard_content / set_clipboard_content - Clipboard access

License

MIT License - see LICENSE file.

Contributing

Contributions welcome! Please feel free to submit issues and pull requests.

Acknowledgments

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