PulseAudio

The PulseAudio backend provides full control over the node’s audio: server info, output selection, global and per-client volume/mute. Works with both PulseAudio and PipeWire (via pipewire-pulse).

Uses a pure-Go PulseAudio native protocol implementation — no libpulse dependency.

Endpoints

Combined state

GET /audio

Returns server info, outputs, and clients in a single response.

Server

GET /audio/server

POST /audio/server/mute
POST /audio/server/volume

Outputs (sinks)

GET /audio/outputs

POST /audio/outputs/{output}/default
POST /audio/outputs/{output}/mute
POST /audio/outputs/{output}/volume

Clients (sink inputs)

GET /audio/clients

POST /audio/clients/{sink}/mute
POST /audio/clients/{sink}/volume

PulseAudio cookie

GET /audio/cookie

Downloads the PulseAudio authentication cookie — useful for setting up network audio sinks.

Events

Event	Trigger
audio.updated	Sink input added or changed
audio.removed	Sink input removed

Configuration

pulseaudio:
  enabled: true
  serve_cookie: true

serve_cookie exposes GET /audio/cookie — save the downloaded cookie on your client as ~/.config/pulse/cookie with 600 permissions to enable network audio streaming.

How it works

The backend connects to PulseAudio via its native protocol over the Unix socket. Real-time events are captured through PulseAudio’s built-in monitoring mechanism.