Ableton Extensions logo

Ableton Extensions

Community
Ronvaknins
ableton-extensions

Scaffold, write, build, and package Ableton Live extensions with the Ableton Extensions SDK (@ableton-extensions/sdk, TypeScript). Use when creating or editing an Ableton Live extension — adding context-menu actions, modal/progress dialogs, or code that manipulates tracks, clips, MIDI notes, devices, tempo, audio rendering, or selections in a Live Set.

Overview

PublisherRonvaknins
Repositoryableton-extensions-skill
Skill nameableton-extensions
Stars
82
Forks
4
Bundled files
3
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by Ronvaknins on GitHub. Read the source before you install it.

Installation

Install the Ableton Extensions AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/Ronvaknins/ableton-extensions-skill/tree/master
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/Ronvaknins/ableton-extensions-skill.git \
  .claude/skills/ableton-extensions
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ableton Extensions in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Ableton Extensions on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Ableton Extensions is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Ableton Extensions SDK

Build tools that run inside Ableton Live. An extension is a Node/TypeScript module bundled to a single CommonJS file that Live's Extension Host loads. It registers commands and wires them to context-menu actions, then reads and mutates the Live Set through a typed object model (tracks, clips, notes, devices, tempo, scenes).

Extensions are an offline editing / automation layer. They are not real-time processors and not devices in the signal chain: they cannot process the live audio or MIDI stream. The SDK has no note-on/off, MIDI-input, or stream callbacks (the only on* callbacks are dialog/registration lifecycle internals) — an extension only acts when the user triggers a command, then mutates the data model. For real-time MIDI/audio transformation (e.g. "play one key, hear a chord"), the right tool is a MIDI/audio effect device in the track chain: Live's built-in MIDI effects (the Chord device does exactly one-note→chord — set Shift voices to e.g. +4, +7) or a Max for Live device — not an Extension. Extensions complement those by transforming clips/notes at edit time, batch-editing the Set, and adding context-menu tools, dialogs, and rendering.

Local SDK location

The SDK distribution lives wherever you unpacked it — referred to below as <sdk-dir>. It contains:

  • docs/ — pre-built HTML documentation (getting-started, essentials, development, design)
  • api/ — TypeDoc HTML API reference
  • examples/ — 7 runnable examples (read these for canonical patterns)
  • ableton-create-extension-<version>.tgz — the scaffolding tool
  • ableton-extensions-cli-<version>.tgz — the build/run/package CLI
  • ableton-extensions-sdk-<version>.tgz — the SDK library

This skill was written against SDK 1.0.0-beta.0 (API version "1.0.0"); newer builds may ship a different <version> — check the actual .tgz filenames in <sdk-dir> and substitute it wherever it appears below. Requires Node.js >= 24.14.1.

Reference files (load as needed)

  • reference/scaffolding.md — create a project, the generated file layout, manifest.json / package.json / build.ts / tsconfig.json, the extensions-cli run|package commands, the .env / EXTENSION_HOST_PATH, debugging, log file locations, packaging to .ablx.
  • reference/api.md — the full object model: ExtensionContext, handles, transactions, Song/Track/Clip/MidiClip/AudioClip/Device/Scene, MIDI NoteDescription, enums (WarpMode, GridQuantization), context-menu scopes, selection types.
  • reference/recipes.md — copy-paste code for every common task: context menus, modal dialogs (with the WebView bridge + Ableton-themed CSS), progress dialogs with cancellation, creating/editing clips and notes, audio import/render, transactions.

The examples/ folder is the source of truth for idiomatic code — prefer reading the closest example to the task over guessing.

The entry point — always this shape

src/extension.ts exports activate. Call initialize once, then register commands and menu actions synchronously:

typescript
import { initialize, type ActivationContext } from "@ableton-extensions/sdk";

export function activate(activation: ActivationContext) {
  const context = initialize(activation, "1.0.0"); // ExtensionContext

  // 1. Register a command (the logic)
  context.commands.registerCommand("myext.doThing", (arg: unknown) => {
    // arg is a Handle, ArrangementSelection, or ClipSlotSelection
    // depending on the scope the menu action was registered for
  });

  // 2. Surface it as a right-click menu item (the trigger)
  context.ui.registerContextMenuAction("AudioClip", "Do Thing", "myext.doThing");
}

context (the ExtensionContext) is the root of everything:

  • context.application.song — the Live Set (tempo, tracks, scenes, …)
  • context.commands — registerCommand(id, cb) / executeCommand(id, …)
  • context.ui — registerContextMenuAction, showModalDialog, withinProgressDialog
  • context.resources — importIntoProject(path), render audio from a track
  • context.environment — storageDirectory, tempDirectory, language
  • context.getObjectFromHandle(handle, Class) — turn a Handle into a typed object
  • context.withinTransaction(() => …) — group mutations into one undo step

Two concepts that trip people up

Handles are not objects. Live passes commands lightweight Handles ({ id: bigint }). Resolve to a typed object with context.getObjectFromHandle(handle, Track). Handles go invalid when the object is deleted, moved in the arrangement, or the Set changes — never cache objects or handles across operations.

Commands receive scope-dependent args. Object scopes ("AudioClip", "MidiTrack", "ClipSlot", "Scene", …) pass a single Handle. Selection scopes pass a structured object: "AudioTrack.ArrangementSelection" / "MidiTrack.ArrangementSelection" pass an ArrangementSelection (time_selection_start, time_selection_end, selected_lanes: Handle[]); "ClipSlotSelection" passes { selected_clip_slots }. Match the scope you register to the arg type you cast.

Standard workflow

  1. Scaffold (new project): npx file:<sdk-dir>/ableton-create-extension-<version>.tgz <dir> — interactive prompts for name, author, Live path, UI support. See reference/scaffolding.md.
  2. Write src/extension.ts — register commands + menu actions; see reference/recipes.md.
  3. Run in Live: npm start (builds with esbuild, then extensions-cli run loads it into the Extension Host). Enable Developer Mode first in Live's Preferences → Extensions.
  4. Debug: console.log/warn/error → ExtensionHost.txt (paths in reference/scaffolding.md). In VS Code, F5 / "Debug Extension".
  5. Package: npm run package → a distributable .ablx archive users drop into Live.

Conventions that matter

  • TypeScript, ESM project, CJS bundle. package.json is "type": "module"; build.ts bundles to format: "cjs", platform: "node". Keep it that way — the Extension Host loads CommonJS.
  • Namespace command IDs ("myext.action") to avoid collisions with other extensions.
  • Most mutations are async (Promise). Batch related edits in context.withinTransaction(() => promises) then await Promise.all(...) so the user gets one undo step and Live updates once.
  • Narrow handle types defensively with instanceof (e.g. resolve to Clip, then check clip instanceof AudioClip) — a menu scope can match objects the code must reject.
  • HTML dialogs are inlined by esbuild (loader: { ".html": "text" }) and passed as a data:text/html, URL — there's no web server.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Ableton Extensions AI skill do?

Scaffold, write, build, and package Ableton Live extensions with the Ableton Extensions SDK (@ableton-extensions/sdk, TypeScript). Use when creating or editing an Ableton Live extension — adding context-menu actions, modal/progress dialogs, or code that manipulates tracks, clips, MIDI notes, devices, tempo, audio rendering, or selections in a Live Set.

Why use Ableton Extensions on TypingMind?

Because you install it once and use it with any model. Ableton Extensions is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Ableton Extensions in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Ronvaknins/ableton-extensions-skill/tree/master. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Ableton Extensions?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Ableton Extensions?

As many as you like. As long as a model supports skills, you can use Ableton Extensions with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Ableton Extensions AI skill free?

It is published on GitHub by Ronvaknins. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇