Live translate

Skill sumitaich1998/jarvisvr/skills/communication/live-translate

An AI agentic operating system for mixed reality on the Meta Quest 3 — your own J.A.R.V.I.S. Multi-agent orchestration, multimodal perception (sight/hearing/gaze), 42 holographic widgets, 20 LLM providers.

Install

npx -y skills add sumitaich1998/jarvisvr --skill live-translate

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Translate a live two-way conversation between two languages in real time, showing both sides on a translator panel. Use for "translate this conversation", "help me talk to them in Spanish", or interpreting back-and-forth speech. Triggers: translate conversation, interpret, talk to them in, live translation, two-way translate, help me communicate.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.8 KB, as published. Nobody here has run it

Live Translate

Run a conversational interpreter: capture each speaker's utterance, translate it, and keep both sides visible on a translator in conversation mode.

Steps

Set the language pair. source_lang/target_lang (BCP-47; auto detects the source). See references/language-codes.md for codes and notes.
Open the translator (show_translator, mode:"conversation", listening:true).
Per utterance: take the perception.audio_scene.ambient_transcript (speaker:"other") or the user's speech, translate_text it, and holo.update the panel's source_text/translated_text. Speak the translation aloud if the user wants voiced interpreting.
Swap direction on swap_languages; toggle capture on toggle_listen.

Output

translator (show_translator, props per registry.json):

{ "widget_type": "translator",
  "transform": { "anchor": "head", "position": [0.0,0.0,1.0], "billboard": true },
  "props": { "source_lang": "es", "target_lang": "en", "mode": "conversation",
             "source_text": "¿Dónde está la estación?", "translated_text": "Where is the station?",
             "listening": true },
  "interactions": ["tap","grab","resize","toggle"] }

agent.speech (voiced interpreting): { "text": "They asked: where is the station?", "final": true }

Bundled resources

references/language-codes.md — BCP-47 quick reference + detection/RTL notes, loaded on demand when resolving an unusual language.

Edge cases

Static sign/menu (not a conversation) → perception's read-and-translate-sign.
Unknown/auto language → set source_lang:"auto"; confirm if detection is low-confidence.
Overlapping speakers → caption the dominant speaker; pair with caption-conversation for full transcript.
Privacy → ambient capture is user-initiated; stop listening (toggle_listen off, perception.request{stream:"ambient_audio", action:"stop"}) when done.
Profanity / sensitive content → translate faithfully and neutrally.