feat(config): add optional Audio config for STT/Music

Add optional Audio *AudioConfig pointer field on Config with STT and
Music sub-structs. Nil-safe — absent in JSON5 decodes as nil, no
breaking change. cfg.Tts retained unchanged for backward compat.
setupAudioExtras stub wired for Phase 3/4 STT/Music provider
registration.

cmd/gateway_agents.go

@@ -274,3 +274,13 @@ func setupTTS(cfg *config.Config) *tts.Manager {
 
 	return mgr
 }
+
+// setupAudioExtras wires optional STT + Music static defaults into the audio
+// Manager. Phase 1 placeholder: body is empty — Phase 3 wires Music providers,
+// Phase 4 wires STT providers. Kept here so the call-site in gateway_setup.go
+// is stable from Phase 1 onward.
+//
+//nolint:unused // Phase 1 stub; wired in Phase 3/4.
+func setupAudioExtras(_ *config.Config, _ *tts.Manager) {
+	// Intentionally empty. Phase 3/4 will register Music/STT providers here.
+}

cmd/gateway_setup.go

@@ -137,11 +137,13 @@ func setupToolRegistry(
 	toolsReg.Register(tools.NewCreateAudioTool(providerRegistry,
 		cfg.Tts.ElevenLabs.APIKey, cfg.Tts.ElevenLabs.BaseURL))
 
-	// TTS (text-to-speech) system — always create TtsTool so config reload can populate it later
+	// Audio system (TTS today, STT/Music/SFX in later phases) — always create
+	// TtsTool so config reload can populate it later.
 	ttsMgr := setupTTS(cfg)
 	if ttsMgr == nil {
 		ttsMgr = tts.NewManager(tts.ManagerConfig{})
 	}
+	setupAudioExtras(cfg, ttsMgr) // Phase 1 stub; Phase 3/4 wire Music/STT.
 	ttsTool = tools.NewTtsTool(ttsMgr)
 	toolsReg.Register(ttsTool)
 	if ttsMgr.HasProviders() {

internal/config/config.go

@@ -50,6 +50,7 @@ type Config struct {
 	Sessions  SessionsConfig  `json:"sessions"`
 	Database  DatabaseConfig  `json:"database"`
 	Tts       TtsConfig       `json:"tts"`
+	Audio     *AudioConfig    `json:"audio,omitempty"` // optional STT/Music defaults (Phase 3/4)
 	Cron      CronConfig      `json:"cron"`
 	Telemetry TelemetryConfig `json:"telemetry"`
 	Tailscale TailscaleConfig `json:"tailscale"`

internal/config/config_audio.go

@@ -0,0 +1,35 @@
+package config
+
+// AudioConfig is the optional cfg.Audio block for STT/Music static defaults.
+// Pointer-typed field on Config so absent/empty JSON silently decodes as nil —
+// existing config.json files that predate Phase 1 continue to load unchanged.
+//
+// Phase 1 ships the shape only (no consumers yet). Phase 3 wires Music; Phase 4
+// wires STT. In both phases the Manager continues to honor per-tenant
+// builtin_tools settings regardless of whether cfg.Audio is set.
+type AudioConfig struct {
+	Stt   *AudioSTTConfig   `json:"stt,omitempty"`
+	Music *AudioMusicConfig `json:"music,omitempty"`
+}
+
+// AudioSTTConfig configures optional static STT defaults. Empty = "no global
+// default; rely on tenant builtin_tools[stt] or per-channel STTProxyURL".
+type AudioSTTConfig struct {
+	Provider   string `json:"provider,omitempty"`    // primary provider name (e.g. "scribe")
+	APIKey     string `json:"api_key,omitempty"`     // may be overridden by env / secrets
+	BaseURL    string `json:"base_url,omitempty"`    // override for enterprise deploys
+	Model      string `json:"model,omitempty"`       // provider-specific
+	Language   string `json:"language,omitempty"`    // BCP-47 hint
+	Fallback   string `json:"fallback,omitempty"`    // provider name to try on primary failure
+	TimeoutMs  int    `json:"timeout_ms,omitempty"`  // default 30000
+}
+
+// AudioMusicConfig configures optional static Music defaults.
+type AudioMusicConfig struct {
+	Provider  string `json:"provider,omitempty"`  // primary provider name (e.g. "elevenlabs")
+	APIKey    string `json:"api_key,omitempty"`
+	BaseURL   string `json:"base_url,omitempty"`
+	Model     string `json:"model,omitempty"`
+	Fallback  string `json:"fallback,omitempty"`
+	TimeoutMs int    `json:"timeout_ms,omitempty"`
+}