You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The core of the framework. Pipeline flow: **Audio In → VAD → STT → LLM → TTS → Audio Out**.
64
64
65
65
Key classes and their roles:
66
+
66
67
-**`Agent`** — Base class holding instructions, tools, and model config. Subclass to override pipeline hooks (`ttsNode`, `realtimeAudioOutputNode`, `sttNode`, `llmNode`). Use `voice.Agent.default.<hook>(this, ...)` to call the base implementation from overrides.
67
68
-**`AgentTask`** — Extends Agent for composable task workflows with isolated chat context and tool scoping.
68
69
-**`AgentSession`** — Orchestrates the full session lifecycle: connects to LiveKit room, manages turn detection, handles interruptions, collects metrics. Entry point: `session.start({ agent, room })`.
@@ -105,6 +106,7 @@ Wire protocol for distributed agents via LiveKit room message channels. `Session
105
106
Each extends `Plugin` base class, auto-registers on import via `Plugin.registerPlugin()`. Pattern: `@livekit/agents-plugin-<provider>`.
0 commit comments