LilArmy LilArmy

Voice Commands

Control LilArmy hands-free with voice commands. Works offline using on-device AI.

How It Works

  1. Press and hold the microphone button in the sidebar
  2. Speak your command clearly
  3. Release to process your voice
  4. The command will be recognized and executed automatically

Voice recognition runs entirely on your device using Whisper.js. Your voice data never leaves your device.

Available Commands

Commands work in both English and Portuguese.

Create a Card

English:

"Create [name]"

Portuguese:

"Criar [nome]"

Example: "Create shopping list" or "Criar lista de compras"

Open a Card

English:

"Open [name]"

Portuguese:

"Abrir [nome]"

Example: "Open meeting notes" or "Abrir notas da reuniao"

Search

English:

"Search [query]"

Portuguese:

"Buscar [termo]" or "Procurar [termo]"

Example: "Search project ideas" or "Buscar ideias de projeto"

Switch View

English:

"Switch to [view]"

Portuguese:

"Mudar para [visao]"

Views: canvas, list, kanban, table, calendar, matrix, chart

Add Tag

English:

"Add tag [name]"

Portuguese:

"Adicionar tag [nome]"

Example: "Add tag urgent" or "Adicionar tag urgente"

Navigation

English:

"Go back" / "Go home"

Portuguese:

"Voltar" / "Inicio"

Dictation Mode

When editing a card, you can dictate text directly. The auto-punctuation feature will automatically add periods, commas, and other punctuation marks.

Voice Punctuation Commands

English:

  • "period" or "full stop" = .
  • "comma" = ,
  • "question mark" = ?
  • "exclamation mark" = !
  • "new line" = new line
  • "new paragraph" = double line break

Portuguese:

  • "ponto" or "ponto final" = .
  • "virgula" = ,
  • "ponto de interrogacao" = ?
  • "ponto de exclamacao" = !
  • "nova linha" = new line
  • "novo paragrafo" = double line break

Stop Dictation

Say "stop dictation" or "parar ditado" to end dictation mode.

Settings

Configure voice recognition in Settings > Voice:

  • Model Size: Choose between Tiny (~40MB, fastest), Base (~75MB, balanced), or Small (~150MB, best accuracy)
  • Language: Auto-detect, English, or Portuguese
  • Parser Mode: Pattern matching (fast, predictable) or AI/LLM (natural language understanding)
  • Auto-Punctuation: Automatically add punctuation during dictation
  • Haptic Feedback: Vibration feedback on mobile devices

Tips

  • Speak clearly and at a moderate pace for best recognition
  • The first use may take a few seconds to load the AI model
  • Larger models (Base, Small) provide better accuracy for accented speech
  • Works completely offline - no internet connection required
  • Use the AI/LLM parser mode for more natural command phrasing
Back to Home