daw-mcp-server
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
daw-mcp-server
如果这个项目对你有用,欢迎点个 Star,让更多做音乐 AI / Agent 的人找到它。
MCP server for DAW automation, rendering and verification — 把 DAW 能调用的接口(FL Studio MIDI scripting / REAPER ReaScript / DAWproject / 文件级工程)封装成 MCP tools,给 训练闭环 和 推理验证 用。
生态(三个仓库)
仓库 | 内容 |
daw-mcp-server(本仓库) | DAW 自动化/渲染/验证 MCP tools(11 个) |
3,713 万 tokens 音乐制作/音频工程文本语料(23 源,含和声/对位/乐器法/总谱分析) | |
鼓 Pattern(10,208)、结构标注(30,336)、DAWproject 规范、参数样本 |
Related MCP server: ComposerBench
能力
Tool | 作用 |
| 服务配置、桥接模式、可用 DAW |
| 各 DAW 可调用 API/命令清单(FL/REAPER/DAWproject/Ardour/PT/Bitwig…) |
| 读 |
| 校验 DAWcoder IR patch(ops 结构 + 参数范围) |
| 把 patch 应用到 |
| 渲染 WAV:配置的 DAW 命令(如 |
| WAV 指标:peak/RMS/LUFS-approx/LRA/频谱质心/频带占比/立体声宽度/相关度 |
| 两个 WAV 的指标差 |
| 闭环:apply → render baseline+patched → analyze → 按 intent 打分 |
| 调用 live DAW 桥(get_state/set_track/set_eq/set_comp/add_marker/transport) |
| 工程师操作/验证结果写入 JSONL(训练数据回流) |
30 秒上手
git clone https://github.com/ShokouP/daw-mcp-server
cd daw-mcp-server
python -m pip install -r requirements.txt
python tests/test_smoke.py # 无需安装 DAW,全部测试应通过
python examples/training_loop.py接入 MCP 客户端(Claude / Codex / DSH):
{
"mcpServers": {
"daw": {
"command": "python",
"args": ["-m", "daw_mcp.server", "--config", "C:/path/to/config.json"]
}
}
}训练脚本直接 import:
from daw_mcp.tools import Context, dispatch
ctx = Context({"telemetry_path": "telemetry.jsonl"})
res = dispatch(ctx, "verify_patch", {"path": "song.rpp", "patch": patch,
"intent": {"rms_dbfs": -1.5}})架构
MCP client (Claude Code / DSH / Codex / Claude Desktop)
│ stdio JSON-RPC (MCP)
▼
daw-mcp-server
├── tools.py tool schemas + dispatch
├── rpp.py .rpp 解析/改写(音轨音量/声像/send/marker)
├── dawproject.py DAWproject ZIP+XML 读取
├── render.py 真实命令渲染 / mock 渲染
├── metrics.py WAV 测量(numpy)
└── bridge.py 队列桥协议(.req.json → .res.json + index.json)
│ 文件队列
▼
adapters/reaper_bridge.lua (REAPER 内 ReaScript 常驻循环)
adapters/flstudio_bridge.py (FL Studio MIDI scripting 脚本)无 DAW 时用 mock 模式(默认):文件级补丁 + 确定性合成渲染 + 指标分析,训练/CI 可完整跑闭环。
配置
{
"render_cmd": "",
"bridge_queue": "C:/daw_mcp_queue",
"bridge_timeout": 15,
"telemetry_path": "daw_telemetry.jsonl",
"mock_tracks": [{"name": "Drums", "volume_db": 0.0, "pan": 0.0}]
}render_cmd为空 → mock 渲染;填reaper -renderproject {project}之类 → 真实渲染。bridge_queue为空 → 内存 mock bridge;填写则启用文件队列桥(REAPER/FL 适配器)。
Live DAW 桥接
REAPER:Actions → Load ReaScript →
adapters/reaper_bridge.lua,改QUEUE_DIR与 serverbridge_queue一致,运行即常驻。FL Studio:
adapters/flstudio_bridge.py复制到...\Image-Line\FL Studio\Settings\Hardware\daw_mcp\device_daw_mcp.py,改QUEUE_DIR,选中该 controller 脚本。细节与注意事项见 adapters/README.md。
测试状态
tests/test_smoke.py:rpp 解析/改写、patch 校验、mock 渲染、WAV 指标、闭环打分、MCP stdio 握手/tools 列表/调用 → ALL TESTS PASSEDexamples/training_loop.py:30 个 patch 的生成-验证-记录闭环;自洽 intent 30/30 通过开发环境未安装 FL Studio / REAPER:live 适配器为模板(协议与 mock 已验证),需在装有 DAW 的机器上加载校对
训练/推理闭环
模型输出 IR patch
→ patch_validate(结构+范围)
→ patch_apply(写 .rpp 副本)
→ render(真实 DAW 命令 或 mock 渲染)
→ analyze(peak/RMS/LUFS-approx/LRA/频谱/立体声)
→ compare / verify_patch(按 intent 打分)
→ bridge_call(通过则应用到 live DAW)
→ telemetry_append(训练数据回流)许可
Apache-2.0(见 LICENSE)。适配器脚本安装到对应 DAW,不包含 DAW 本体或受版权手册。
This server cannot be deployed
Maintenance
Related MCP Connectors
Compose, mix, master and export in Logic Pro, REAPER or Ableton Live by talking to an AI engineer.
Browser DAW for AI assistants: build multitrack songs with MIDI, effects, stems and generation.
Autonomous music production for AI agents with MIDI generation, QC and provenance.
Score video & make music as an agent: 50+ AI instruments play your MIDI, audio, stems, Logic/Ableton
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables AI-powered music production by integrating with REAPER DAW via MCP protocol, allowing users to control project settings, MIDI editing, and VST plugins through natural language.10-
- FlicenseNot gradedqualityBmaintenanceProvides AI agents with reversible DAW control for music production, enabling inspection of instruments and parameters, batch edits, and evaluated revision runs with audio feedback and rollback.2-
- AlicenseBqualityCmaintenanceEnables MCP-capable AI agents to control REAPER through typed tools and a Lua escape hatch for project creation, track management, recording and editing, MIDI writing, mixing, rendering, and batch operations.501MIT
- AlicenseNot gradedqualityBmaintenanceEnables an AI agent to compose and mix directly inside FL Studio: writing and clearing notes, setting tempo, levels and plugin parameters, inspecting what FL currently holds, checking which channels sound, and rendering the arrangement to WAV.3MIT