Agent stays "speaking" forever when the avatar participant leaves mid-reply
维护者通常 1 天内回复
评估
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 新手友好度
- 54/100
- Issue 类型
- 缺陷
- 描述清晰度
- 基本清楚
- 活跃度
- 活跃
- 技术栈
- python
调研方向
Start by locating DataStreamAudioOutput and its playout wait and clear-buffer timeout handling, then compare the Synthesia plugin's participant-disconnect teardown. Trace how participant_disconnected should release a pending segment, including after the avatar rejoins; done means pending playout completes as interrupted and the agent no longer stays in speaking.
由索引模型根据 Issue 内容生成。
描述
Title: Agent stays "speaking" forever when the avatar participant leaves mid-reply
With any avatar built on DataStreamAudioOutput, if the avatar participant disconnects while a reply is pending, lk.playback_finished never arrives. wait_for_playout() has no timeout, so the agent stays speaking until the user interrupts. With allow_interruptions=False, or a user waiting in silence, it never recovers.
It also stays stuck after the avatar rejoins with the same identity: the old segment is never finished, so every later playout wait is behind it.
Repro
- Start an
AgentSessionwith an avatar plugin (e.g. Anam). session.say("<long text>", allow_interruptions=False)- Mid-reply, remove the avatar participant. Killing the avatar worker also works once the server's departure timeout fires.
- The speech never completes;
agent_statestaysspeaking.
DataStreamAudioOutput doesn't listen for participant_disconnected at all. Only the Synthesia plugin handles avatar loss, and its teardown doesn't release pending playout either.
Expected: pending playout completes as interrupted, the same as the existing clear-buffer timeout fallback.
- 主要语言
- Python
- 星标
- 14.6k
- 派生
- 3.9k
- 平均合并
- 3 天 11 小时
- 30 天内合并 PR
- 196
环境准备
这个项目没有提供开发容器、Dockerfile 或贡献指南，环境需要你自己搭建：先看它的 README，通用步骤见我们的新手贡献指南。
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
livekit/agents 的其他 Issue
-
Assistant markup is stripped from chat history when expressive mode is off可能已有人在做 @ShivamSharma43 于 17 天前认领。 未关闭
难度 2/5 1-3 小时 新手友好度 72/100
维护者通常 1 天内回复
-
docs: openai.LLM vs openai.responses.LLM for HTTP OpenAI-compat hosts可能重新可做 @AbarnaaSree 于 56 天前认领，目前没有进行中的 PR。 未关闭
难度 2/5 1-3 小时 新手友好度 72/100
维护者通常 1 天内回复
-
维护者通常 1 天内回复
-
TranslationRoomIO: unset_track() before a track is linked stalls startup, so the translation is never published可能已有人在做 @RaphaelFakhri 今天认领。 未关闭
维护者通常 1 天内回复
-
Python AgentServer health check (GET /) returns 200 before the worker has registered with LiveKit可能已有人在做 @ninglinLiu 于 1 天前认领。 未关闭
难度 3/5 1-2 天 新手友好度 35/100
维护者通常 1 天内回复
相似的 Issue
-
难度 1/5 1 小时以内 新手友好度 72/100
slavakurilyak/awesome-ai-agents#845 ·
维护者通常 1 天内回复
-
tests: pin that the atomic writers in utils/paths.py stage the temporary file in the target's directory可能已有人在做 关联的 PR 仍在进行中或已合并。 未关闭difficulty:2 enhancement good first issue
难度 1/5 1-3 小时 新手友好度 92/100
MakazhanAlpamys/Soup#2178 ·
维护者通常 1 天内回复
-
[BUG] dcg.waggr permutation p-values are folded twice, so near-zero scores get the smallest p-values未关闭
难度 2/5 1-3 小时 新手友好度 75/100
scverse/rapids-singlecell#854 ·
维护者通常 1 天内回复
-
Slang input: int8_t4_packed and uint8_t4_packed are copied into CrossGL可能已有人在做 @ninglinLiu 今天认领。 未关闭area:frontend backend:slang bug direction:input
难度 2/5 1-3 小时 新手友好度 70/100
维护者通常 1 天内回复
-
random_elements(unique=True) on an OrderedDict: ZeroDivisionError once only zero weights remain, AssertionError with use_weighting=False可能已有人在做 关联的 PR 仍在进行中或已合并。 未关闭
难度 2/5 1-3 小时 新手友好度 85/100
维护者通常 13 天内回复