cleanshot-mcp
Enables AI assistants to control CleanShot X for screenshots, recordings, OCR, and annotations via natural language commands.
README
CleanShot MCP Server (Unofficial)
Control CleanShot X from any MCP-compatible AI assistant with simple commands like "take a screenshot" or "capture this area".
Built with the Model Context Protocol (MCP) - a standard for connecting AI assistants to external tools. Works with Amp, Claude Desktop, and other MCP-compatible applications.
Requirements
- macOS
- CleanShot X installed and running
- ⚙️ Settings > Advanced > API > ☑️ Allow applications to control CleanShot X
- Node.js 18+
Quick Start
1. Install the MCP Server
# Option 1: Use without installing (recommended)
npx cleanshot-mcp
# Option 2: Install globally
npm install -g cleanshot-mcp
2. Configure Your MCP Client
Add this to your MCP client configuration:
{
"amp.mcpServers": {
"cleanshot": {
"command": "npx",
"args": ["cleanshot-mcp"],
"env": {}
}
}
}
3. Start Using It
Simply tell your AI assistant:
- "Take a fullscreen screenshot and copy it"
- "Capture this area: x: 100, y: 100, width: 500, height: 300"
- "Open CleanShot settings"
Examples
Natural Language Commands:
| Command | What It Does |
|---|---|
| "Take a fullscreen screenshot and copy it" | Captures entire screen and copies to clipboard |
| "Capture this area: x: 100, y: 100, width: 500, height: 300" | Captures specific screen region |
| "Open CleanShot settings" | Opens CleanShot preferences |
| "Extract text from this area" | Uses OCR to extract text from screen region |
Available Tools
This MCP server provides 17 CleanShot tools including:
- Screenshots: Area capture, fullscreen, window capture, self-timer
- Recording: Screen recording with custom areas
- Text Extraction: OCR from any screen region
- Annotation: Open annotation tools for images
- Management: History, settings, desktop icon control
🔧 View Complete Tool List | 📕 API Reference
Development
Building
npm run build
Development Mode
npm run dev
Testing
Make sure CleanShot is installed and running, then test individual commands:
# Test basic functionality
node dist/index.js
Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Submit a pull request
Support
- Issues with this MCP server: Open a GitHub issue
- MCP protocol questions: MCP Documentation
[!IMPORTANT] Please DO NOT contact CleanShot support for problems with this MCP server. This is an unofficial integration created by fans of their product.
⭐ Star this repo if you find it useful!
Created with Amp by Sourcegraph
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。