Repository navigation
feat(voice): hold a key and speak orders - #3
Open
leorivastech wants to merge 2 commits into
Open
leorivastech wants to merge 2 commits into
leorivastech wants to merge 2 commits into
Conversation
github-actions
Bot
added
vouch:trusted
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Voice orders let users hold a key and control T3 Code without moving between the keyboard and chat. Right Ctrl is the default, with a per-device key setting. Navigation and scrolling execute after pauses while the key stays held; other orders wait for release.
Demo (56s, English): t3-voice-orders.mp4
https://github.com/leorivastech/t3code/releases/download/voice-orders-demo/t3-voice-orders.mp4
Changing chats
Hold Right Ctrl and say the chat. The highlight moves, and the key stays down.
Scrolling
Same hold. “Go to the top”, then “Down”, then “Up”. The numbered list is the proof: items 1–9, then 8–22, then back to 1–9.
• Live navigation & scrolling: Jump to numbered threads, move next/previous, and scroll repeatedly while holding the key.
• Composer & actions: Write a draft, send, stop or settle, select a model or effort, and combine navigation/new-thread/model/message actions.
• Desktop screen capture: Capture the desktop screen and attach the last capture to a draft.
• Local & cloud: Use Groq or OpenAI for transcription and interpretation, or the included optional local Whisper helper for navigation/scrolling without any cloud request.
• Privacy & hygiene: Store the voice key in the environment secret store and redact it for clients. Cancel recordings on blur, another key, hidden window, or the 60-second limit.
This PR is against my fork and contains only voice changes. Mods and Telegram are excluded.
Try it
Use Node 24 and Vite+ (
vp). From a clone of this fork:Add a project, connect an agent provider, and open Settings → Integrations → Voice. Save a Groq/OpenAI key for all orders, or turn on Local Whisper for offline navigation/scrolling.
Validation
• 113 focused TypeScript tests and 5 Python helper tests pass.
• Scoped typechecks passed for contracts, client-runtime, server, web and desktop.
• Changed-file lint passes; server/web/desktop builds passed.
• The demo above is a real UI recording: “Two”, “Next” and “Four” change the open chat while Right Ctrl stays held, and “Go to the top”, “Down” and “Up” move the list.