Skip to content
AgentPrism

VoiceConversationProtocol

Namespace AgentPrism · Assembly AgentPrism.Core.dll

The frame names and the audio formats of the conversation WebSocket.

public static class VoiceConversationProtocol

Inheritance

Section titled “Inheritance”

object ← VoiceConversationProtocol

Inherited Members

Section titled “Inherited Members”

object.GetType(), object.MemberwiseClone(), object.ToString(), object.Equals(object?), object.Equals(object?, object?), object.ReferenceEquals(object?, object?), object.GetHashCode()

Remarks

Section titled “Remarks”

The names stay aligned with the existing SSE contract; the user interface does not carry two different mental models. These are a stable contract: to change them breaks clients.

Text frames are JSON and binary frames are raw audio. The direction follows from the name of the frame, not from its type.

Fields

Section titled “Fields”

ClientCancel

Section titled “ ClientCancel”

Interruption (barge-in): stop the generation.

public const string ClientCancel = "cancel"

Field Value

Section titled “Field Value”

string

ClientCommit

Section titled “ ClientCommit”

Speech ended; answer now.

public const string ClientCommit = "commit"

Field Value

Section titled “Field Value”

string

ClientStart

Section titled “ ClientStart”

Starts the conversation.

public const string ClientStart = "start"

Field Value

Section titled “Field Value”

string

ClientStop

Section titled “ ClientStop”

Closes the connection.

public const string ClientStop = "stop"

Field Value

Section titled “Field Value”

string

ServerAudioEnd

Section titled “ ServerAudioEnd”

The audio chunks ended.

public const string ServerAudioEnd = "audioEnd"

Field Value

Section titled “Field Value”

string

ServerAudioStart

Section titled “ ServerAudioStart”

The binary frames that follow are audio chunks.

public const string ServerAudioStart = "audioStart"

Field Value

Section titled “Field Value”

string

ServerDone

Section titled “ ServerDone”

The turn ended.

public const string ServerDone = "done"

Field Value

Section titled “Field Value”

string

ServerError

Section titled “ ServerError”

An error.

public const string ServerError = "error"

Field Value

Section titled “Field Value”

string

ServerReady

Section titled “ ServerReady”

The server is ready for the conversation.

public const string ServerReady = "ready"

Field Value

Section titled “Field Value”

string

ServerRunStarted

Section titled “ ServerRunStarted”

The run started; the frame carries its identifier.

public const string ServerRunStarted = "runStarted"

Field Value

Section titled “Field Value”

string

ServerText

Section titled “ ServerText”

A text response (caption).

public const string ServerText = "text"

Field Value

Section titled “Field Value”

string

ServerTranscript

Section titled “ ServerTranscript”

The transcribed speech text.

public const string ServerTranscript = "transcript"

Field Value

Section titled “Field Value”

string

SubProtocol

Section titled “ SubProtocol”

The name of the WebSocket sub-protocol. The handshake echoes it back.

public const string SubProtocol = "agentprism.voice.v1"

Field Value

Section titled “Field Value”

string

TokenSubProtocolPrefix

Section titled “ TokenSubProtocolPrefix”

The sub-protocol prefix that carries the bearer token.

public const string TokenSubProtocolPrefix = "agentprism.token."

Field Value

Section titled “Field Value”

string

Remarks

Section titled “Remarks”

The token is not put in the query string: the address is written to server logs, to reverse proxy logs and to the browser history. A browser cannot add a custom header to a WebSocket handshake; the standard way out is the Sec-WebSocket-Protocol header.