| name | voice-audio-mode |
| description | Manage voice mode and audio mode for response style. Voice mode triggers shorter, conversational replies. Audio mode triggers TTS audio file responses. Use when the user says voice mode on/off, audio mode on/off, on the go, on mobile, on my AirPods, back at desk, mobile mode, desk mode, or text mode. |
| tier | REACTOR |
| triggers | ["voice mode on","voice mode off","audio mode on","audio mode off","on the go","I'm on the go","on mobile","on my AirPods","back at my desk","back at desk","mobile mode","desk mode","text mode"] |
voice-audio-mode — Response Style Toggle
Switches Jarvis between voice-optimized and full-detail response modes based on your context.
Modes
Voice Mode (on the go)
- Responses are short and conversational — 1-3 sentences max
- No markdown, no lists, no tables in spoken output
- Long results go to Discord text channel, TL;DR spoken
- Jarvis assumes you're mobile, hands-free, can't look at a screen
- Trigger: "on the go", "I'm mobile", "voice mode on", "on my AirPods"
Text Mode (at desk)
- Full responses, full detail, markdown OK
- Longer analysis and structured output welcome
- Jarvis assumes you have a screen in front of you
- Trigger: "back at my desk", "text mode", "voice mode off", "desk mode"
How to Use
Just say the trigger phrase naturally in conversation:
"Jarvis, I'm heading out — on the go mode."
"Back at my desk."
"Voice mode on."
"Switch to text mode."
What Changes
| Behavior | Voice Mode | Text Mode |
|---|
| Response length | 1-3 sentences | Full detail |
| Format | Plain speech | Markdown OK |
| Long results | TL;DR spoken + Discord | Full in chat |
| Sub-agent output | Spoken summary | Full report |
Implementation
When voice mode is toggled on:
- Set a session flag
VOICE_MODE=true
- Apply jarvis-voice-briefing pattern to all subsequent responses
- Acknowledge: "Voice mode on. I'll keep it brief."
When voice mode is toggled off:
- Clear the session flag
- Acknowledge: "Text mode. Full detail available."
This mode is session-scoped — it resets when the conversation ends. For persistent mode across sessions, store the preference in hAIveMind.
