Member Junction
    Preparing search index...

    Resolved configuration for LiveKitNativeMeetingSdk. Credentials resolve upstream (MJ credential system / provider Configuration) — this object carries already-resolved values; never inline secrets at a call site.

    interface LiveKitNativeSdkConfig {
        AccessToken?: string;
        AgentVision?: boolean;
        ApiKey?: string;
        ApiSecret?: string;
        BotDisplayName?: string;
        InboundSampleRate?: number;
        InboundVideoRate?: number;
        InboundVideoStreams?: number;
        NativeModuleSpecifier?: string;
        OutboundSampleRate?: number;
        ScreenIn?: boolean;
        Url?: string;
        VideoCameraMaxDimension?: number;
        VideoIn?: boolean;
        VideoJpegQuality?: number;
        VideoScreenMaxDimension?: number;
        VideoSpeakerHoldMs?: number;
        VideoSpeakerOnsetMs?: number;
    }
    Index

    Properties

    AccessToken?: string

    A pre-signed LiveKit access token, when the deployment mints it out-of-band instead of via key/secret.

    AgentVision?: boolean

    Whether the agent watches the meeting: set by the room coordinator when the agent's setting is on and its session takes video. Only then does the room client read anyone's camera or screen (and only people who allow it).

    ApiKey?: string

    The LiveKit API key (resolved upstream). Used to mint the join token when not pre-signed.

    ApiSecret?: string

    The LiveKit API secret (resolved upstream). Used only to sign the join token.

    BotDisplayName?: string

    The bot's display name (defaults applied by the bridge when absent).

    InboundSampleRate?: number

    PCM rate (Hz) the agent's realtime model consumes — the rate inbound room audio is resampled to before reaching the model. Threaded from the model via the engine (IRealtimeSession.InputSampleRate). Default 24000 (OpenAI); Gemini Live = 16000. Mismatch = the agent never hears the user on the bridge.

    InboundVideoRate?: number

    The session's inbound video rate (frames per second), threaded from the model by the engine.

    InboundVideoStreams?: number

    How many inbound video streams the session takes, threaded from the model by the engine. 0 or absent: none.

    NativeModuleSpecifier?: string

    The module specifier of the native LiveKit room-client wrapper to load (e.g. an internal wrapper package name or an absolute path to a sidecar entry). Required in production; tests inject a loader.

    OutboundSampleRate?: number

    PCM rate (Hz) the agent's model emits — the bot's published voice track rate. Default 24000.

    ScreenIn?: boolean

    Whether the provider allows shared screens in (its ScreenIn flag), passed down by the bridge.

    Url?: string

    The LiveKit room server ws URL (resolved upstream).

    VideoCameraMaxDimension?: number

    Override of the camera frame size cap (longer side, px).

    VideoIn?: boolean

    Whether the provider allows camera video in (its VideoIn flag), passed down by the bridge.

    VideoJpegQuality?: number

    Override of the JPEG quality (1-100).

    VideoScreenMaxDimension?: number

    Override of the screen frame size cap (longer side, px).

    VideoSpeakerHoldMs?: number

    Override of how long (ms) a camera stays in view, from its first frame, before another camera may replace it.

    VideoSpeakerOnsetMs?: number

    Override of how long (ms) a person must lead the active-speaker list before the view moves to their camera.