All pull requests

Pull requests

• fix(voice): snapshot manually committed user turns

  #2969·rosetta-livekit-bot[bot] opened Nov 10, 2026Bot·

  ·
  5

• fix(voice): stop transcript words of a cleared segment after resume

  #2968·Brian-1812 opened Nov 10, 2026·

  ·
  1

  1

• fix(voice): reject handoffs to running agents

  #2966·rosetta-livekit-bot[bot] opened Nov 10, 2026Bot·

  ·
  1

• feat(deepgram): Support usage tags on v1 STT

  #2965·snydermb opened Nov 9, 2026·6 of 8 tasks·

  ·
  7

• chore: remove speechStreamTestState test-only helper from meta STT

  #2964·detail-app[bot] opened Nov 9, 2026Bot·6 of 7 tasks·

  ·
  1

• feat(examples): add the four homepage demo agents

  #2963·rosetta-livekit-bot[bot] opened Nov 9, 2026Bot·

  ·
  7

• fix(voice): don't log a tool's StopResponse as an error

  #2961·Brian-1812 opened Nov 8, 2026·

  ·
  1

• fix(ipc): report failed job entrypoints as failed

  #2959·rosetta-livekit-bot[bot] opened Nov 8, 2026Bot·

  ·
  4

• feat(vartaloop): add @livekit/agents-plugin-vartaloop TTS plugin

  #2958·roshanpuru opened Nov 8, 2026·

  ·
  14

• fix(google): count Gemini thinking tokens in LLM usage

  #2956·pcbeingused333 opened Nov 7, 2026·4 of 8 tasks·

  ·
  1

• feat(simulation): add SimulationContext.defines()

  #2955·u9g opened Nov 6, 2026Contributor·

  ·
  1

• docs(inference): Deprecate provider pinning

  #2954·shawnfeldman opened Nov 6, 2026Contributor·3 of 4 tasks·

  ·
  1

• docs(agents): sync XaiSTTModels API report for xai/stt-2

  #2953·detail-app[bot] opened Nov 6, 2026Bot·5 of 8 tasks·

  ·
  1

  

• fix(soniox): keep sending text pushed after a TTS retry

  #2952·anzemur opened Nov 6, 2026Contributor·

  ·
  4

• fix(soniox): send every audio frame to the new socket after an STT reconnect

  #2951·anzemur opened Nov 6, 2026Contributor·

  ·
  1

• feat(simulation): add simulation.choose() to compare the options of a knob

  #2950·u9g opened Nov 6, 2026Contributor·

  ·
  1

• feat(soniox): add agent text input and endpoint sensitivity

  #2949·anzemur opened Nov 6, 2026Contributor·

  ·
  1

• feat(inference): Add rime/prism to the realtime models

  #2947·rosetta-livekit-bot[bot] opened Nov 5, 2026Bot·

  ·
  1

• feat(amd): tag AMD's gateway requests with a client feature header

  review_effort:medium
  #2945·chenghao-mou opened Nov 5, 2026Member·

  ·
  1

• fix(gpt-live): cap caller audio queued before session.started

  #2944·rosetta-livekit-bot[bot] opened Nov 5, 2026Bot·

  ·
  5

• feat(kalpa): add Kalpa TTS plugin

  #2943·rosetta-livekit-bot[bot] opened Nov 5, 2026Bot·

  ·
  5

• fix(voice): only the job's room shares the job-owned session host

  #2942·u9g opened Nov 5, 2026Contributor·

  ·
  4

• feat(fishaudio): add realtime STT

  #2940·rosetta-livekit-bot[bot] opened Nov 4, 2026Bot·

  ·
  6

• fix(voice): send per-reply instructions as their own system message

  #2938·himanshumehta opened Nov 4, 2026·

  ·
  4

• fix(avatar): finish pending playout when the avatar participant leaves

  #2937·himanshumehta opened Nov 4, 2026·

  ·
  1

  11