ableton-live-mcp

ableton-live-mcp

Local MCP server for inspecting and controlling Ableton Live through a local HTTP bridge. Enables LLMs to perform production workflows like MIDI import, track editing, mixing, mastering, and export.

Category
访问服务器

README

Ableton Live MCP

Local MCP stdio server for inspecting and controlling Ableton Live through a local bridge.

Current package version: 0.1.0.

This repo does not assume a specific Ableton integration layer. The MCP server speaks JSON-RPC over stdio to an MCP client, then forwards tool calls to a local HTTP bridge that you can implement with Max for Live, Ableton's Extensions SDK, a Python Remote Script helper, or another Live automation layer.

Product Workflows

Start with docs/product-workflows.md for end-to-end setup, MIDI import, instrument assignment, arrangement, mixing, mastering, validation, and export workflows. The guide also documents safe operating rules for open user sets, destructive commands, explicit export approval, and unsupported bridge operations.

Tools

Read and inspect:

  • ableton_get_status
  • ableton_get_project
  • ableton_get_arrangement
  • ableton_list_plugins
  • ableton_search_browser
  • ableton_diagnose_plugins
  • ableton_get_production_report
  • ableton_diagnose_playback
  • ableton_get_meters
  • ableton_list_returns
  • ableton_list_buses
  • ableton_get_clip_notes
  • ableton_launch_clip
  • ableton_launch_scene
  • ableton_get_device_parameters
  • ableton_analyze_audio
  • ableton_analyze_mix
  • ableton_get_bridge_observability
  • ableton_evaluate_action_risk
  • ableton_list_risk_policy
  • ableton_match_preset_intent
  • ableton_list_workflow_plans
  • ableton_get_workflow_plan

Project, transport, track, and clip editing:

  • ableton_create_snapshot
  • ableton_rollback_snapshot
  • ableton_set_tempo
  • ableton_save_project
  • ableton_set_signature
  • ableton_start_transport
  • ableton_stop_transport
  • ableton_create_midi_track
  • ableton_duplicate_track
  • ableton_freeze_track
  • ableton_flatten_track
  • ableton_modify_track
  • ableton_create_midi_clip
  • ableton_consolidate_clip
  • ableton_delete_clip
  • ableton_humanize_clip
  • ableton_quantize_clip
  • ableton_apply_groove
  • ableton_import_midi

Devices, routing, returns, arrangement, mastering, and export:

  • ableton_load_device
  • ableton_load_master_device
  • ableton_select_vst_for_midi
  • ableton_set_device_parameter
  • ableton_reorder_device
  • ableton_delete_device
  • ableton_set_automation
  • ableton_create_return_track
  • ableton_modify_return
  • ableton_delete_return_track
  • ableton_modify_master
  • ableton_apply_mastering_chain
  • ableton_analyze_and_apply_mastering
  • ableton_export_render
  • ableton_bounce_tracks
  • ableton_insert_arrangement_clip
  • ableton_add_locator

Production Workflow

For meaningful LLM control, the recommended sequence is:

  1. ableton_get_status, ableton_get_project, ableton_get_arrangement, and ableton_get_production_report to inspect transport, tracks, devices, clips, locators, routing, meters, returns, and mixer state.
  2. ableton_create_snapshot before broad edits. For real user sets, also save or duplicate the .als file before destructive, export, or mastering work.
  3. ableton_list_plugins and ableton_search_browser so the LLM can choose available instruments, effects, presets, samples, racks, grooves, plugins, and Max devices. Use ableton_match_preset_intent first when the request is musical, such as realistic violin, concert piano, classical guitar, flute, or concert hall mastering.
  4. ableton_import_midi, ableton_create_midi_clip, or both.
  5. ableton_humanize_clip, ableton_quantize_clip, or ableton_apply_groove for controlled MIDI timing edits.
  6. ableton_select_vst_for_midi, ableton_load_device, or ableton_load_master_device to assign sounds and processing.
  7. ableton_duplicate_track, ableton_consolidate_clip, ableton_insert_arrangement_clip, ableton_add_locator, ableton_modify_track, ableton_modify_return, ableton_modify_master, ableton_set_device_parameter, and ableton_set_automation for arrangement and mix moves.
  8. ableton_freeze_track and ableton_flatten_track when committing tracks to audio is supported by the active bridge.
  9. ableton_apply_mastering_chain for master bus preparation.
  10. ableton_export_render, ableton_bounce_tracks, and ableton_analyze_audio, and ableton_analyze_mix for approved export, stem, and audio-validation workflows. File analysis runs locally in the MCP process and does not depend on Ableton's Remote Script meters.

Risk tiers and default blocking recommendations are documented in docs/risk-policy.md and exposed through ableton_evaluate_action_risk and ableton_list_risk_policy. High-level plan-only workflows are exposed through ableton_list_workflow_plans and ableton_get_workflow_plan. Render/export support is documented in docs/render-export-plan.md; the current Python Remote Script bridge returns explicit 501 unsupported responses for render-backed routes instead of simulating Live renders.

Quick Start

From npm:

npx -y @jterrats/ableton-live-mcp --help

Install the bundled Ableton Remote Script:

npx -y @jterrats/ableton-live-mcp install-remote-script --app-path "/Applications/Ableton Live 12 Lite.app"

Then restart Ableton Live, select AbletonMcpBridge in Preferences -> Link, Tempo & MIDI, and verify the installation:

npx -y @jterrats/ableton-live-mcp doctor --app-path "/Applications/Ableton Live 12 Lite.app"

Run the MCP stdio server:

npx -y @jterrats/ableton-live-mcp

Local development from this repo:

npm run smoke

Run the server:

npm start

Run the deterministic local development bridge:

npm run bridge

For dry-run development without Ableton:

ABLETON_MCP_DRY_RUN=1 npm start

MCP Client Config

Use this shape in an MCP client that supports stdio servers:

{
  "mcpServers": {
    "ableton-live": {
      "command": "npx",
      "args": ["-y", "@jterrats/ableton-live-mcp"],
      "env": {
        "ABLETON_BRIDGE_URL": "http://127.0.0.1:9789"
      }
    }
  }
}

Set ABLETON_MCP_DRY_RUN=1 in env when you want the tools to return the intended bridge calls without touching Ableton.

Bridge

The local bridge must listen on ABLETON_BRIDGE_URL and implement the HTTP contract in docs/ableton-bridge-contract.md.

This repo includes a deterministic development bridge for local MCP integration tests. It listens on 127.0.0.1:9789 by default, or on ABLETON_BRIDGE_PORT when configured, and implements the product workflow surface without requiring Ableton Live.

For the real Ableton Live adapter, install the Python Remote Script documented in docs/ableton-python-remote-script.md. That script runs inside Ableton and exposes the same bridge endpoints on 127.0.0.1:9789.

On this machine Ableton Live was detected at /Applications/Ableton Live 12 Lite.app. Install the bundled Remote Script with:

npx -y @jterrats/ableton-live-mcp install-remote-script --app-path "/Applications/Ableton Live 12 Lite.app"

If macOS rejects writes to /Applications with Operation not permitted, close Ableton and retry with sudo -E npx -y @jterrats/ableton-live-mcp install-remote-script --app-path "/Applications/Ableton Live 12 Lite.app". If sudo is still blocked, grant App Management or Full Disk Access to your terminal app in System Settings -> Privacy & Security, then rerun the installer. Finder manual copy with authentication is also valid.

Release readiness checks are documented in docs/release-checklist.md.

Keep the bridge local-only. Do not bind it to a public network interface unless you add authentication and understand the risk of remote DAW control.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选