Accept inbound RTVI speaking signals as turn frames - #5715
Open
kixelated wants to merge 2 commits into
Open
Conversation
A cooperating client that runs VAD at the microphone can now send user-started-speaking / user-stopped-speaking (and the raw vad-user-* pair) client-to-server. Turn messages become proposals so ExternalUserTurnStrategies owns barge-in; VAD messages become VAD frames so the default VAD start strategy works without a server analyzer. Co-Authored-By: grok-4.6 <noreply@x.ai>
Co-Authored-By: grok-4.6 <noreply@x.ai>
kixelated
marked this pull request as ready for review
September 12, 2026 06:47
Codecov Report✅ All modified and coverable lines are covered by tests.
🚀 New features to boost your workflow:
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
RTVI speaking messages (
user-started-speaking/user-stopped-speakingand the rawvad-user-*pair) were server-to-client only. A cooperating client that runs VAD at the microphone could not drive the bot's turn boundaries.This makes those types inbound as well:
user-started-speaking/user-stopped-speakingbecomeProposedUserStartedSpeakingFrame/ProposedUserStoppedSpeakingFrame, soExternalUserTurnStrategiesowns barge-in instead of adopting an already-announced turn (which would skip the interruption).vad-user-started-speaking/vad-user-stopped-speakingbecomeVADUserStartedSpeakingFrame/VADUserStoppedSpeakingFrame, so the default VAD start strategy works without a server-side analyzer.Unknown types still fail loud. Server-to-client emission is unchanged.
Test plan
tests/test_rtvi_processor.py::TestRTVIInboundSpeakingcovers the four mappings and that_handle_messagedoes not send an unsupported-type error.(written by grok-4.6)