mixxx-mcp
MCP server for Mixxx DJ software that enables AI agents to control transport, mixing, EQ, loops, hotcues, and effects in real time through MIDI and OSC.
README
mixxx-mcp
MCP (Model Context Protocol) server for Mixxx DJ Software. Lets AI agents (Claude, Claude Code, etc.) control Mixxx in real time — transport, mixing, EQ, loops, hotcues, effects.
Architecture
AI Agent (Claude)
│
▼ MCP tools
mixxx-mcp Python Server
│ ┌─────────────────────┐
├─ MIDI CC out ───────▶│ Virtual MIDI port │
│ │ "mixxx-mcp" │
│ └────────┬────────────┘
│ ▼ CC → engine.setValue()
│ ┌────────────────────┐
│ │ Mixxx + mixxx-mcp │
│ │ .js controller │
│ │ script │
◀── OSC UDP ──────────┤ broadcasts state │
OscStateStore └────────────────────┘
Write path: mcp.tool → MidiBridge → rtmidi CC → Mixxx JS script → engine.setValue()
Read path: Mixxx JS → OSC UDP → OscStateStore → mcp.tool → AI
Setup
1. Install Python server
pip install -e ".[dev]"
# or without dev deps:
pip install mcp python-osc python-rtmidi
2. Install Mixxx controller script
Copy both files to your Mixxx controllers directory:
# macOS
cp mixxx-mcp.js mixxx-mcp.midi.xml \
~/Library/Containers/org.mixxx.mixxx/Data/Library/Application\ Support/Mixxx/controllers/
# Linux
cp mixxx-mcp.js mixxx-mcp.midi.xml ~/.mixxx/controllers/
# Windows
cp mixxx-mcp.js mixxx-mcp.midi.xml %LOCALAPPDATA%\Mixxx\controllers\
3. Enable in Mixxx
- Open Mixxx → Preferences → Controllers
- Select mixxx-mcp from the device list
- Check Enable
- Click OK
Note (Windows): Virtual MIDI ports require loopMIDI. Create a port named
mixxx-mcpbefore starting the server.
4. Run the MCP server
# stdio mode (Claude Desktop / claude-code)
python main.py
# HTTP mode (remote / multi-client)
python main.py --http --port 8080
5. Configure Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"mixxx": {
"command": "python",
"args": ["/path/to/mixxx-mcp/main.py"],
"env": {}
}
}
}
Available Tools
| Tool | Description |
|---|---|
play(deck) |
Start playback |
stop(deck) |
Stop playback |
cue(deck) |
Trigger CUE |
sync(deck) |
Toggle BPM sync |
set_volume(deck, value) |
Channel fader 0.0–1.0 |
set_crossfader(value) |
Crossfader -1.0–1.0 |
set_eq(deck, low, mid, high) |
EQ bands 0.0–4.0 |
set_pregain(deck, value) |
Trim 0.0–4.0 |
set_rate(deck, value) |
Tempo pitch -1.0–1.0 |
nudge_tempo(deck, direction, size) |
Tempo nudge up/down |
set_loop(deck, beats) |
Activate beat loop |
exit_loop(deck) |
Deactivate loop |
halve_loop(deck) / double_loop(deck) |
Resize loop |
set_hotcue(deck, slot) |
Set hotcue 1–8 |
goto_hotcue(deck, slot) |
Jump to hotcue |
clear_hotcue(deck, slot) |
Delete hotcue |
beatjump(deck, beats) |
Jump ±N beats |
get_deck_state(deck) |
Read live deck state |
get_mixer_state() |
Read master mixer state |
get_all_state() |
Full OSC state dump |
send_control(group, key, value) |
Raw control escape hatch |
toggle_effect(unit, effect, enabled) |
Effect on/off |
set_effect_mix(unit, value) |
Effect wet/dry |
OSC State
Mixxx broadcasts control changes via the JS script's engine.makeConnection() callbacks.
The Python server listens on UDP 57121 at path /mixxx/<group>/<key>.
To fully enable OSC push (vs poll), the JS callbacks in mixxx-mcp.js need to forward
values via a UDP socket. This requires a Mixxx build with liblo, or you can extend
the JS script to use XMLHttpRequest to POST state to a local HTTP endpoint.
Mixxx Controls Reference
Full control list: https://manual.mixxx.org/latest/en/chapters/appendix/mixxx_controls.html
Key groups:
[Channel1]–[Channel4]— Decks[Master]— Mixer master[Sampler1]–[Sampler64]— Samplers[EffectRack1_EffectUnit1]–[_EffectUnit4]— Effect units[EffectRack1_EffectUnit1_Effect1]–[_Effect3]— Individual effects[Playlist]— Library navigation
License
MIT
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。