speak

Original：🇺🇸 English
Translated
1 scriptsChecked / no sensitive code detected

Text-to-speech using edge-tts or macOS say. Use when user says "speak", "say", "read aloud", or wants text spoken.

2installs
Sourcesoul-brews-studio/oracle-skills-cli
Added on

NPX Install

npx skill4agent add soul-brews-studio/oracle-skills-cli speak

Tags

Translated version includes tags in frontmatter
text-to-speechedge-ttsmacos-sayvoice-synthesiscommand-line-tool

SKILL.md Content

View Translation Comparison →

/speak - Text-to-Speech

Speak text using edge-tts (high-quality neural voices) with fallback to macOS 
say
.

Quick Start

bash
/speak "Hello world"                    # Default English voice
/speak --thai "สวัสดีครับ"                # Thai voice
/speak --voice "en-GB-RyanNeural" "Hi"  # Specific voice
/speak --mac "Hello"                    # Force macOS say

Options

OptionDescription
--thai
Use Thai voice (th-TH-NiwatNeural)
--female
Use female voice
--voice NAME
Specific edge-tts voice
--mac
Force macOS say command
--rate RATE
Speech rate (edge-tts: +/-50%, mac: 100-300)
--list
List available voices

Default Voices

LanguageVoice
Englishen-US-GuyNeural (male)
Englishen-US-JennyNeural (female)
Thaith-TH-NiwatNeural (male)
Thaith-TH-PremwadeeNeural (female)

Scripts

Located in 
src/skills/speak/scripts/
:
ScriptPurpose
speak.ts
Main TTS script

Run Script

bash
cd src/skills/speak/scripts
bun speak.ts "Text to speak"
bun speak.ts --thai "ภาษาไทย"
bun speak.ts --list

Requirements

  • edge-tts: 
    pip install edge-tts
    (optional, for high-quality voices)
  • macOS say: Built-in (fallback)

How It Works

  1. Check if edge-tts is available
  2. If yes: Use edge-tts with neural voice
  3. If no: Fall back to macOS 
    say
    command
  4. Play audio through default output

Examples

bash
# Speak Gemini response
/speak "The answer is 42"

# Thai greeting
/speak --thai "สวัสดีครับ ยินดีต้อนรับ"

# British accent
/speak --voice "en-GB-RyanNeural" "Brilliant!"

# List all voices
/speak --list