REAPER MCP AI Analyzing

English | 中文

English

A Model Context Protocol (MCP) server for REAPER DAW with AI-powered audio analysis. Control REAPER through natural language and get intelligent mix suggestions from AI models with audio capabilities.

🚀 Features

• 25+ MCP Tools for REAPER control (tracks, FX, routing, media items)
• AI Audio Analysis - Analyze mixes with ANY audio-capable AI model
• File-based IPC - No network required, works on macOS/Windows/Linux
• Multi-Provider Support - OpenAI, Google Gemini, Anthropic Claude, and any OpenAI-compatible API

🤖 Supported AI Models

| Provider | Supported Models | Audio Input | |----------|------------------|-------------| | OpenAI | gpt-4o-audio-preview, gpt-4o-mini-audio-preview | ✅ Native | | Google | gemini-2.5-pro, gemini-2.5-flash | ✅ Native | | Anthropic | claude-4-opus, claude-4-sonnet | ✅ File attachment | | Custom | Any OpenAI-compatible API | Depends on model |

📁 Project Structure

reaper-mcp-ai-analyzing/
├── src/
│   ├── index.ts              # MCP Server main entry
│   ├── types/
│   │   └── reaper.ts         # Type definitions
│   ├── utils/
│   │   ├── env.ts            # Environment config
│   │   ├── logger.ts         # Logging utility
│   │   └── audio.ts          # Audio processing
│   ├── bridge/
│   │   └── file-client.ts    # File IPC client
│   └── ai/
│       ├── analyzer.ts       # Audio analysis coordinator
│       └── providers/
│           ├── base.ts       # Base AI provider
│           └── openai.ts     # OpenAI implementation
├── bridge/
│   └── file-bridge.lua       # REAPER Lua bridge script
└── package.json

🔧 Installation

# Clone the repository
git clone https://github.com/YangJue213-spec/reaper-mcp-ai-analyzing.git
cd reaper-mcp-ai-analyzing

# Install dependencies
npm install

# Build the project
npm run build

# Configure environment
cp .env.example .env
# Edit .env with your OpenAI API key

⚙️ Configuration (.env)

# ============================================
# AI Provider Configuration
# ============================================

# Provider Type: 'openai', 'gemini', 'anthropic', or 'custom'
AI_PROVIDER=openai

# --- OpenAI Configuration ---
OPENAI_API_KEY=sk-your-openai-key-here
OPENAI_BASE_URL=https://api.openai.com/v1
AUDIO_MODEL_NAME=gpt-4o-audio-preview
# Alternative models: gpt-4o-mini-audio-preview

# --- Google Gemini Configuration ---
# GEMINI_API_KEY=your-gemini-key-here
# GEMINI_BASE_URL=https://generativelanguage.googleapis.com/v1beta
# AUDIO_MODEL_NAME=gemini-2.5-pro
# Alternative models: gemini-2.5-flash

# --- Anthropic Claude Configuration ---
# ANTHROPIC_API_KEY=your-anthropic-key-here
# ANTHROPIC_BASE_URL=https://api.anthropic.com
# AUDIO_MODEL_NAME=claude-4-opus-20250514
# Alternative models: claude-4-sonnet-20250514

# ============================================
# REAPER Configuration
# ============================================
REAPER_SCRIPT_TIMEOUT=30000

🎮 MCP Tools

Project & Track Control

• get_project_info - Get project sample rate, tempo, track count
• get_track_info - Get track details (volume, pan, FX count, etc.)
• create_track - Create new track with optional name
• delete_track - Delete track by index
• set_track_name - Rename a track
• set_track_volume - Set volume in dB
• set_track_pan - Set pan (-1 to 1)
• set_track_send - Create send to another track
• set_track_output - Route track output

FX Control

• get_track_fx - List all FX on a track
• add_fx_to_track - Add plugin (supports Waves, FabFilter, generic)
• remove_fx_from_track - Remove FX by index
• get_fx_params - Get all parameters of an FX
• set_fx_param - Set parameter by absolute value
• set_fx_param_normalized - Set parameter (0-1 normalized)
• set_fx_enabled - Enable/disable FX

Media Items

• get_item_info - Get item position, length, fades
• split_item - Split item at position

AI Analysis (Core Feature)

• analyze_and_suggest_mix - Render and analyze audio with AI
• start_audio_analysis - Start async analysis task
• get_analysis_status - Check analysis progress

🔌 MCP Configuration

Add to your MCP client settings (Claude Desktop, Cline, etc.):

{
  "mcpServers": {
    "reaper-ai": {
      "command": "node",
      "args": ["/path/to/reaper-mcp-ai-analyzing/build/index.js"],
      "env": {
        "OPENAI_API_KEY": "your_api_key_here",
        "OPENAI_BASE_URL": "https://api.openai.com/v1",
        "AUDIO_MODEL_NAME": "gpt-4o-audio-preview"
      }
    }
  }
}

🎵 How to Use

1. Setup REAPER Bridge

1. Open REAPER
2. Go to Actions → Show Action List
3. Click New Action → Load ReaScript
4. Select bridge/file-bridge.lua
5. The bridge is now running (check REAPER console for "REAPER MCP Bridge started")

2. Start MCP Server

npm start

3. Use Natural Language Prompts

Example 1: Analyze a track

"Analyze the lead vocal track (track 0) and suggest EQ improvements"

Example 2: Check mix loudness

"Check the overall mix loudness and tell me if it meets streaming standards"

Example 3: Add FX and analyze

"Add a compressor to the bass track, then analyze if the dynamics are better"

📝 Example Prompts

1. "Get project info and list all tracks"
2. "Set track 0 volume to -6 dB and pan to 0.2"
3. "Add Pro-Q 3 to track 1 and analyze the frequency balance"
4. "Check REAPER connection status"
5. "Analyze the master mix from 0 to 30 seconds"

中文

带有 AI 音频分析功能 的 REAPER MCP 服务器。通过自然语言控制 REAPER，并从任何支持音频输入的 AI 模型获取智能混音建议。

🚀 功能特性

• 25+ MCP 工具 用于 REAPER 控制
• AI 音频分析 - 使用任何支持音频的 AI 模型分析混音
• 文件 IPC - 无需网络，支持 macOS/Windows/Linux
• 多提供商支持 - OpenAI、Google Gemini、Anthropic Claude

🤖 支持的 AI 模型

| 提供商 | 支持的模型 | 音频输入 | |--------|-----------|---------| | OpenAI | gpt-4o-audio-preview, gpt-4o-mini-audio-preview | ✅ 原生支持 | | Google | gemini-2.5-pro, gemini-2.5-flash | ✅ 原生支持 | | Anthropic | claude-4-opus, claude-4-sonnet | ✅ 文件附件 | | 自定义 | 任何兼容 OpenAI 的 API | 取决于模型 |

🔧 安装

git clone https://github.com/YangJue213-spec/reaper-mcp-ai-analyzing.git
cd reaper-mcp-ai-analyzing
npm install
npm run build
cp .env.example .env
# 编辑 .env 配置你的 AI 提供商

⚙️ 配置 (.env)

# ============================================
# AI 提供商配置
# ============================================

# 提供商类型: 'openai', 'gemini', 'anthropic', 或 'custom'
AI_PROVIDER=openai

# --- OpenAI 配置 ---
OPENAI_API_KEY=sk-your-openai-key-here
OPENAI_BASE_URL=https://api.openai.com/v1
AUDIO_MODEL_NAME=gpt-4o-audio-preview
# 备选模型: gpt-4o-mini-audio-preview

# --- Google Gemini 配置 ---
# GEMINI_API_KEY=your-gemini-key-here
# GEMINI_BASE_URL=https://generativelanguage.googleapis.com/v1beta
# AUDIO_MODEL_NAME=gemini-2.5-pro
# 备选模型: gemini-2.5-flash

# --- Anthropic Claude 配置 ---
# ANTHROPIC_API_KEY=your-anthropic-key-here
# ANTHROPIC_BASE_URL=https://api.anthropic.com
# AUDIO_MODEL_NAME=claude-4-opus-20250514
# 备选模型: claude-4-sonnet-20250514

# ============================================
# REAPER 配置
# ============================================
REAPER_SCRIPT_TIMEOUT=30000

🔌 MCP 配置

添加到 Claude Desktop、Cline 等 MCP 客户端：

{
  "mcpServers": {
    "reaper-ai": {
      "command": "node",
      "args": ["/path/to/reaper-mcp-ai-analyzing/build/index.js"],
      "env": {
        "AI_PROVIDER": "openai",
        "OPENAI_API_KEY": "your_api_key_here",
        "AUDIO_MODEL_NAME": "gpt-4o-audio-preview"
      }
    }
  }
}

🎵 使用方法

1. 设置 REAPER 桥接

1. 打开 REAPER
2. 进入 动作 → 显示动作列表
3. 点击 新建动作 → 加载 ReaScript
4. 选择 bridge/file-bridge.lua
5. 桥接已启动（在 REAPER 控制台查看）

2. 启动 MCP 服务器

npm start

3. 使用自然语言

示例 1：分析轨道

"分析主唱轨道（轨道 0）并建议 EQ 调整"

示例 2：检查响度

"检查整体混音响度，是否符合流媒体标准"

📝 提示词示例

1. "获取工程信息并列出所有轨道"
2. "将轨道 0 音量设为 -6 dB，声像设为 0.2"
3. "给轨道 1 添加 Pro-Q 3 并分析频响平衡"
4. "检查 REAPER 连接状态"
5. "分析主混音从 0 到 30 秒的部分"

🛠️ 故障排除

"REAPER file bridge is not available"

• 确保 Lua 脚本已在 REAPER 中运行
• 检查 REAPER 控制台是否有错误信息
• 确认 /tmp/reaper-mcp 目录可访问

"AI Analyzer not initialized"

• 检查 .env 文件中的 OPENAI_API_KEY
• 确保已运行 npm run build

📜 License

MIT License - See LICENSE file