All pull requests

Pull requests

• fix(voice): drain unconsumed VAD/STT input branches in AudioRecognition

  #2765·edycutjong opened Oct 18, 2026·5 of 8 tasks·

  ·
  1

  2

• perf(cartesia): emit 20 ms frames so the first audio leaves sooner

  #2763·bounjee opened Oct 17, 2026·6 of 8 tasks·

  ·
  6

• perf(agents): concatenate audio buffers with TypedArray.set

  #2762·bounjee opened Oct 17, 2026·6 of 8 tasks·

  ·
  2

• fix(elevenlabs): redial a Scribe realtime socket that stops answering

  #2759·cosmin-staicu opened Oct 17, 2026·

  ·
  5

• fix(elevenlabs): let a stalled multi-stream connection be retired

  #2758·cosmin-staicu opened Oct 17, 2026·

  ·
  5

• fix(deepgram): recover Nova streaming STT from a dead socket without losing speech

  #2757·cosmin-staicu opened Oct 17, 2026·

  ·
  5

• perf(agents): read room audio input in 20 ms frames without a frame processor

  #2756·cosmin-staicu opened Oct 17, 2026·

  ·
  1

• fix(cartesia): detect a silently dropped STT socket instead of hanging

  #2755·cosmin-staicu opened Oct 17, 2026·

  ·
  5

• feat(deepgram): support speed parameter in Flux TTS

  #2754·rosetta-livekit-bot[bot] opened Oct 17, 2026Bot·

  ·
  4

• fix(google): hold a turn's transcript that arrives during a blocking tool call

  #2752·anzemur opened Oct 16, 2026Contributor·

  ·
  7

• fix(gradium): identify the STT plugin to the API with x-api-source

  #2750·rosetta-livekit-bot[bot] opened Oct 16, 2026Bot·

  ·
  6

• feat: add MCP refresh and result handling parity

  review_effort:high
  #2748·chenghao-mou opened Oct 16, 2026Member·2/2·8 of 10 tasks·

  ·
  1

• fix: preserve async tool completion after progress

  review_effort:medium
  #2747·chenghao-mou opened Oct 16, 2026Member·1/2·7 of 9 tasks·

  ·
  1

• feat(room_io): warn when agent and client both apply noise cancellation

  #2745·lukasIO opened Oct 16, 2026Contributor·

  ·
  3

• fix(cartesia): remove sunset sonic models, add sonic-3.6

  #2743·cartesia-amarat opened Oct 16, 2026·5 of 8 tasks·

  ·
  2

• Version Packages

  #2741·github-actions[bot] opened Oct 15, 2026Bot·

  ·

• fix(workflows): support turnHandling in warm transfers

  #2737·rosetta-livekit-bot[bot] opened Oct 15, 2026Bot·

  ·
  1

• feat(agents): add DecisionModel for background decisions

  #2735·rosetta-livekit-bot[bot] opened Oct 15, 2026Bot·

  ·
  6

• fix(elevenlabs): place text-to-dialogue character times on the context clock

  #2734·OliverA100 opened Oct 14, 2026·6 of 8 tasks·

  ·
  1

  7

• fix(llm): treat null for an optional tool argument as omitted

  #2732·u9g opened Oct 14, 2026Contributor·

  ·
  3

• feat(aitwin): add Ai Twin plugin for browser speakText()

  #2730·Swaraj07082 opened Oct 14, 2026·8 tasks·

  ·
  6

• fix(anthropic): drop unpaired tool calls and lead user turns with tool_result

  #2729·Apeksha9389 opened Oct 14, 2026·6 of 7 tasks·

  ·
  1

  2

• fix(audio): stop audioFramesFromFile writes from rejecting unhandled after the consumer stops

  #2728·Apeksha9389 opened Oct 14, 2026·6 of 7 tasks·

  ·
  1

  6

• fix(voice): remove waitForOrAbort's abort listener once the race settles

  #2727·Apeksha9389 opened Oct 14, 2026·6 of 7 tasks·

  ·
  1

  1

• fix(soniox): send finalize when the STT stream is flushed

  #2725·MeerHusam opened Oct 14, 2026·

  ·
  1

  7