SuperCmd

Raycast + Wispr Flow + Speechify + Memory + AI

Open-source launcher for macOS with Raycast-compatible extensions, voice workflows, and AI-native actions.

Open-source launcher for macOS: Raycast + Wispr Flow + Speechify + Memory + AI in one app.

SuperCmd gives you Raycast-style extension workflows, hold-to-speak dictation, natural text-to-speech, and AI actions backed by configurable providers and memory.

What It Is

SuperCmd is an Electron + React launcher that focuses on Raycast extension compatibility while remaining community-driven and open source.

Key Features

• Raycast extension compatibility runtime (@raycast/api and @raycast/utils shims)
• Extension discovery/install/update flow
• Script command support
• Wispr Flow-style voice input: hold to speak, release to type
• Speechify-style read-aloud flow for selected text
• Memory-aware AI support (Supermemory integration)
• AI integrations (Ollama / OpenAI / Anthropic / ElevenLabs for speech)
• Native macOS helpers (hotkeys, color picker, speech, snippet expansion)

Tech Stack

• Electron (main process)
• React + Vite (renderer)
• TypeScript
• Swift binaries for macOS-native integrations

Project Structure

src/main/        Electron main process, IPC, extension execution, settings
src/renderer/    React UI + Raycast compatibility layer
src/native/      Swift native helpers
extensions/      Installed/managed extension data
dist/            Build output

How To Run

Prerequisites

• macOS (required for native Swift modules)
• Node.js 18+
• npm
• Xcode Command Line Tools (swiftc available)

Install

npm install

Development

npm run dev

This starts:

• TypeScript watch for main process
• Vite dev server for renderer
• Electron app in development mode

Build

npm run build

Package App

npm run package

Output artifacts are generated under out/.

AI + Memory Setup (Detailed)

You can configure everything from the app UI:

1. Launch SuperCmd.
2. Open Settings.
3. Go to AI tab.
4. Turn AI on (enabled = true in settings).
5. Pick your default provider (OpenAI, Claude, or Ollama).
6. Add the required key(s) and model settings.

Provider keys

• OpenAI: set openaiApiKey
• Anthropic (Claude): set anthropicApiKey
• Ollama: set ollamaBaseUrl (default http://localhost:11434)

Speech / voice-related keys

• ElevenLabs (for STT/TTS options): set elevenlabsApiKey
• Built-in Edge TTS option does not require an API key

ElevenLabs quick voice test

Working preset voices in SuperCmd:

• Rachel (21m00Tcm4TlvDq8ikWAM)
• Domi (AZnzlk1XvdvUeBnXmlld)
• Bella (EXAVITQu4vr4xnSDxMaL)
• Antoni (ErXwobaYiN019PkySvjV)
• Elli (MF3mGyEYCl7XYWbV9V6O)
• Josh (TxGEqnHWrfWFTfGW9XjX)
• Arnold (VR6AewLTigWG4xSOukaG)
• Adam (pNInz6obpgDQGcFmaJgB)
• Sam (yoZ06aMxZJJ28mfd3POQ)

Sample curl command:

curl -X POST "https://api.elevenlabs.io/v1/text-to-speech/21m00Tcm4TlvDq8ikWAM?output_format=mp3_44100_128" \
  -H "xi-api-key: $ELEVENLABS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text":"Hello from SuperCmd","model_id":"eleven_multilingual_v2"}' \
  --output elevenlabs-sample.mp3

Memory keys

• Supermemory API key: supermemoryApiKey
• Supermemory client id: supermemoryClient
• Supermemory base URL: supermemoryBaseUrl (default https://api.supermemory.ai)
• Optional local mode: supermemoryLocalMode

Where settings are stored

All app settings (including AI keys) are persisted in:

~/Library/Application Support/SuperCmd/settings.json

Important fields inside settings.json:

{
  "ai": {
    "enabled": true,
    "provider": "openai",
    "openaiApiKey": "",
    "anthropicApiKey": "",
    "elevenlabsApiKey": "",
    "supermemoryApiKey": "",
    "supermemoryClient": "",
    "supermemoryBaseUrl": "https://api.supermemory.ai",
    "supermemoryLocalMode": false,
    "ollamaBaseUrl": "http://localhost:11434",
    "defaultModel": "openai-gpt-4o-mini",
    "speechToTextModel": "native",
    "textToSpeechModel": "edge-tts"
  }
}

Optional environment variable fallbacks

• ELEVENLABS_API_KEY
• SUPERMEMORY_API_KEY
• SUPERMEMORY_CLIENT
• SUPERMEMORY_BASE_URL
• SUPERMEMORY_LOCAL

Contributing

1. Fork the repository and create a feature branch.
2. Keep compatibility with existing Raycast extensions.
3. Follow current code organization:
   • API/runtime work in src/renderer/src/raycast-api/
   • System-level logic in src/main/
   • UI views/hooks in src/renderer/src/
4. Run build checks before opening a PR:

npm run build

5. In your PR, include:
   • What changed
   • Why it changed
   • Compatibility impact (if any)
   • How you tested it

What's Pending

Current pending/roadmap items:

• Complete OAuth implementation in the Raycast API compatibility layer
• Expand BrowserExtension support beyond current stubs
• Improve selected text/Finder item behavior with macOS permissions handling
• Harden asset loading/path resolution for extension assets
• Add a broader automated test suite (unit + integration + compatibility)
• Continue parity work for edge-case Raycast APIs and behaviors

Useful Commands

npm run dev            # Start local development
npm run build          # Build main, renderer, and native modules
npm run build:main     # Build Electron main process TS
npm run build:renderer # Build renderer with Vite
npm run build:native   # Compile Swift helpers
npm run package        # Build and package app with electron-builder

References

• Raycast API docs: https://developers.raycast.com/api-reference/
• Raycast extension store: https://www.raycast.com/store

GitHub Star History