Skip to content

Configuration Reference

Config File Location

~/.claude/cc-vox.toml

Created automatically on first run. Editable with any text editor or via slash commands.

Full Default Config

[core]
enabled = true
voice = "af_heart"       # see voices below
backend = "auto"         # auto | kokoro | fish-speech | pocket-tts | chatterbox | qwen3-tts

[tuning]
speed = 1.0              # 0.5-2.0 (kokoro only)
max_sentences = 2        # max sentences in spoken summary (1-10)
fallback = true          # try other backends when forced one is down

[style]
prompt = ""

# ── Available Voices ──────────────────────────────
# af_heart (default)  F  American   alias: alba
# af_bella            F  American   alias: azure
# af_nicole           F  American   alias: fantine
# af_sarah            F  American   alias: cosette
# af_sky              F  American   alias: eponine
# am_adam             M  American   alias: marius
# am_michael          M  American   alias: jean
# bf_emma             F  British    alias: azelma
# bm_george           M  British

Settings

[core]

enabled

Type: bool | Default: true

Enable or disable voice feedback globally. When disabled, all hooks become no-ops and the say script exits immediately.

enabled = false

voice

Type: string | Default: "af_heart"

Voice name for TTS. Accepts both Kokoro names (e.g., af_bella) and pocket-tts aliases (e.g., azure). See Voices for the full catalog.

voice = "am_adam"

backend

Type: string | Default: "auto" Valid values: auto, kokoro, fish-speech, pocket-tts, chatterbox, qwen3-tts

TTS backend preference. In auto mode, cc-vox tries backends in priority order (Fish Speech -> Chatterbox -> Qwen3-TTS -> Kokoro -> pocket-tts).

backend = "kokoro"

[tuning]

speed

Type: float | Default: 1.0 | Range: 0.5 -- 2.0

Speech speed multiplier. Only supported by the Kokoro backend; other backends ignore this value. Clamped to the valid range on read.

speed = 1.3

max_sentences

Type: int | Default: 2 | Range: 1 -- 10

Maximum sentences in the spoken summary. Affects both the 📢 marker reminder and the truncation fallback.

max_sentences = 4

fallback

Type: bool | Default: true

When a forced backend is unavailable, fall through to auto-detection. If false, TTS fails silently when the forced backend is down.

fallback = false

[style]

prompt

Type: string | Default: ""

Custom instruction for voice summary personality. Injected into the system prompt and the headless Claude fallback.

prompt = "be upbeat and encouraging"

[internal]

just_disabled

Type: bool | Default: false

Internal flag set when the user runs /voice:speak stop. Consumed once by the UserPromptSubmit hook to inject a "voice disabled" message, then cleared. Not intended for manual editing.