MCP MIDI Bridge
An Electron desktop application that bridges LLM-driven music generation with DAWs by converting NoteSequence JSON from AI models into MIDI data that can be played, recorded, and manipulated in any digital audio workstation.
README
MCP MIDI Bridge
Overview
MCP MIDI Bridge is an Electron-based desktop application that acts as a bridge between LLM-driven music generation (via the Model Context Protocol, MCP) and any DAW (Digital Audio Workstation) that accepts MIDI input. It enables AI-generated or AI-edited musical content (in Magenta NoteSequence JSON format) to be easily played, recorded, and manipulated in a DAW via a virtual MIDI device.
This project has been recently refactored to use a modern technology stack, including Next.js for the user interface and TypeScript for the entire codebase.
Features
- MCP API Server: An Express-based server that receives and updates musical content (NoteSequence JSON) from LLMs via HTTP.
- Virtual MIDI Output: Creates a virtual MIDI device that any DAW can connect to, powered by
easymidi. - Multi-Channel Support: Full support for all 16 MIDI channels with General MIDI instruments.
- Configurable Port: Run multiple instances on different ports for parallel workflows.
- Song Cache: Stores song data for persistence between sessions.
- User Dashboard: A modern, responsive UI built with Next.js and React, for viewing and playing songs.
- MIDI Import/Export: Support for importing and exporting MIDI files (coming soon).
- TypeScript Codebase: The entire project is written in TypeScript for improved type safety and maintainability.
Technology Stack
- Electron: Cross-platform desktop application framework.
- Next.js: React framework for building the user interface.
- TypeScript: Superset of JavaScript that adds static types.
- Express: Web framework for creating the MCP API server.
- EasyMIDI: Library for creating virtual MIDI devices.
- Tailwind CSS: Utility-first CSS framework for styling the UI.
- Magenta: Python library for music generation (used in the Python backend).
Development
Prerequisites
- Node.js 18+
- Python 3.8+ (for Magenta features)
- npm or yarn
Setup
-
Clone the repository:
git clone https://github.com/your-username/mcp-midi.git cd mcp-midi -
Install dependencies:
npm install -
Start the development server: This command will concurrently start the Next.js development server and the Electron application.
npm run dev
Build
To build the application for production, run the following command:
npm run build
This will create a distributable package in the dist directory.
License
Apache 2.0
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。