Documentation
¶
Overview ¶
Package voiceagent provides an embeddable Voice Agent service.
Voice Agent is the realtime audio-to-audio mode: a duplex WebSocket session with the underlying live model (Gemini Live, OpenAI Realtime, or a pipeline fallback) where the user and the agent take audio turns in sequence. Use this when the host needs brainstorming, support, or follow-up dialogue rather than a one-shot result.
For tool registration, lifecycle hooks, and session memory, build on top of this with github.com/kombifyio/SpeechKit/pkg/speechkit/agentkit.
Construct an instance with [New], passing a provider and the strict- mode policy fields from the host config.
Index ¶
Constants ¶
This section is empty.
Variables ¶
View Source
var ErrMissingProvider = errors.New("speechkit voiceagent: provider is required")
Functions ¶
This section is empty.
Types ¶
type Service ¶
type Service struct {
// contains filtered or unexported fields
}
func NewService ¶
func (*Service) CurrentSession ¶
Click to show internal directories.
Click to hide internal directories.