Skip to content
Back to skills

Voice Studio

ASecurity

Use when the user requests spoken voice responses, audio generation, speech synthesis, speaking directly to the user, or expressive speech playback through VoiceStudio

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 21, 2026
ai-agentsbashreactexpress

Security analysis

A96/100
  • mediumUses curl or wget to download content

Pro scans all 2 files and shows the line behind each finding

Scanned September 21, 2026

npx -y skills add LFTPadilla/agent-dev-kit --skill voice-studio --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Voice Studio?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Voice Studio
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/lftpadilla-voice-studio/badge)](https://www.skillsdirectory.com/skills/lftpadilla-voice-studio)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: voice-studio
description: Use when the user requests spoken voice responses, audio generation, speech synthesis, speaking directly to the user, or expressive speech playback through VoiceStudio
---

# VoiceStudio

Synthesize spoken audio responses and play expressive speech locally via GPU.

## Usage

```bash
# Speak in background (non-blocking, default: explainer voice, 1.1x speed, 24 steps)
vsay -b "Task completed successfully."

# Expressive speech with pauses and non-verbal tokens
vsay -b "[surprise-wa] Incredible! The benchmark finished... [breathing] and ran fast. [laughter]"
```

## Flags

- `-b, --bg`: Play in background without blocking caller.
- `-v <voice>`: `explainer` (default), `companion`, `enthusiast`, `narrator`, `radio`, `deep`, or profile ID.
- `-s <float>`: Speed multiplier (default: `1.1`).
- `-N <int>`: Diffusion steps (default: `24`; `16` fast, `32` studio).
- `--cfg <float>`: Guidance scale (default: `2.0`).
- `-o <path>`: Save WAV to target file.
- `--no-play`: Generate audio without playing.
- `-l <code>`: Language code (default: `es`).
- `-L, --list`: List installed voice profiles.

## Expressive Tokens & Prosody

Embed bracketed tokens for vocal inflections:
- **Reactions:** `[laughter]`, `[surprise-wa]`, `[surprise-ah]`, `[sigh]`, `[crying]`
- **Hesitations & Feedback:** `[uhm]`, `[confirmation-en]`, `[dissatisfaction-hnn]`, `[shh]`, `[cough]`
- **Questions:** `[question-en]`, `[question-ah]`, `[question-oh]`, `[question-ei]`, `[question-yi]`

**Pacing rules:**
1. Combine `...` and `[breathing]` between clauses for natural breathing pauses.
2. Use punctuation marks (!, ?, ¡, ¿) to shape expressive pitch contours.
3. Place commas before conjunctions to induce realistic pitch dips.

## Health Check

```bash
curl -s http://127.0.0.1:3900/health
# Verify: "device": "cuda (AMD Radeon Graphics)"
```

Files in this skill

  • SKILL.md1.8 KB
  • scripts/vsay.sh3.4 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…