Tools

voice_speak

Synthesize speech from text.

Parameter	Default	Description
text	(required)	Text to synthesize
voice	—	Voice ID or preset name
speed	1.0	Speed multiplier (0.5–2.0)
mood	—	Humor mood: dry, roast, chaotic, cheeky, cynic, zoomer
format	"wav"	Output format: wav, mp3, ogg, raw
artifactMode	"path"	Delivery mode: path or base64
outputDir	—	Subdirectory within output root
sfx	false	Enable [ding], [chime] etc.

voice_dialogue

Multi-speaker dialogue synthesis using Speaker: line format.

Parameter	Default	Description
script	(required)	Dialogue in Speaker: line format
cast	—	Speaker-to-voice mapping, e.g. { "Alice": "af_sky" }
speed	1.0	Speed multiplier
concat	true	Combine all lines into a single file
debug	false	Include cue_sheet in response
artifactMode	"path"	Delivery mode: path or base64
outputDir	—	Subdirectory within output root

When no cast is provided, speakers are auto-assigned voices from the approved list.

voice_status

Returns engine health, available voices, presets, and backend info. Takes no arguments.

Use this to check what voices are available, which backend is active, and whether the engine is healthy.

voice_interrupt

Stop or rollback active synthesis.

Parameter	Default	Description
streamId	—	Specific stream to interrupt
reason	—	Why: user_spoke, context_change, timeout, manual

voice_inner_monologue

Ephemeral micro-utterances for ambient narration. Requires --ambient flag or VOICE_SOUNDBOARD_AMBIENT_ENABLED=1.

Parameter	Default	Description
text	(required)	Max 500 chars, auto-redacted
category	"general"	One of: general, thinking, observation, debug