agents
agents copied to clipboard
livekit
Build real-time multimodal AI applications 🤖🎙️📹
### Feature Type Would make my life easier ### Feature Description Hello LiveKit Team, I'd like to request support for the Polish language in the Turn Detector model I'm currently...
### Bug Description I just upgraded livekit agents from 1.2.18 to 1.3.3. Now, the resume false interruption behavior is completely broken. This is the behavior when, if the user does...
add a `STTCapabilities.flush` to indicate if the stt supports flush (manual commit), and make `stt.StreamAdapter` work with streaming STT. use cases: 1. only send audio frames to STT when VAD...
### Feature Type Would make my life easier ### Feature Description Sometime the agent interrupts the user. This is because while the user finished speaking a first time. He then...
## Description Cartesia connection in LiveKit Agents closes routinely during STT processing, see attached stack trace. Frequency is ~10%. Please investigate. ## Error Details When using Cartesia, the connection sometimes...
### Bug Description I have a time-consuming process in one of my `function_tool`'s. This process appends utterances to `ChatContext` via `say` and `generate_reply` I have noticed that the following scenario...
### Bug Description Our agent was running stable in production for weeks, then suddenly all job requests started failing with DuplexClosed errors. Redeploying the exact same code without any changes...
Inconsistent behavior between RealtimeSession version and AudioRecognition version of generate_reply
### Bug Description Currently, the generate_reply function processes `instructions` differently for the realtime vs audio recognition version implementations. The audio recognition version appends the instructions, whereas the realtime api replaces...
### Bug Description Sometimes google return no candidate and fallback mechanisim not working how to fix? ### Expected Behavior If gemini response has no candidate it should fallback to other...
As discussed in #3918 some STTs may misfire creating false speech recognition. To fix this @chenghao-mou suggested rewriting `StreamAdapter` to be able to work with stream capable STT's and only...
