Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
alsk1992 avatar

Voice

  • 24 installs
  • 610 repo stars
  • Updated June 26, 2026
  • alsk1992/cloddsbot

voice is a skill that adds wake-word detection, speech-to-text, and text-to-speech voice control to an agent.

About

voice adds a voice interface to an agent with wake-word detection, speech-to-text, and text-to-speech. A developer uses it to control an agent by voice, transcribe speech, and speak responses back. It supports multiple STT providers (Whisper, Vosk, Google) and TTS providers (ElevenLabs, macOS say, espeak).

  • Wake-word detection with speech-to-text and text-to-speech
  • Pluggable STT (Whisper, Vosk, Google) and TTS (ElevenLabs, macOS say, espeak)
  • Voice-controlled commands for a trading agent

Voice by the numbers

  • 24 all-time installs (skills.sh)
  • Ranked #9,912 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
At a glance

voice capabilities & compatibility

Capabilities
transcription
Works with
openai
Use cases
transcription
Platforms
macOS
Pricing
Bring your own API key
From the docs

What voice says it does

Voice-controlled interface with wake word detection, speech-to-text, and text-to-speech.
SKILL.md
provider: 'whisper', // 'whisper' | 'vosk' | 'google'
SKILL.md
npx skills add https://github.com/alsk1992/cloddsbot --skill voice

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs24
repo stars610
Last updatedJune 26, 2026
Repositoryalsk1992/cloddsbot

What it does

Add wake-word voice control, transcription, and speech output to an agent.

Who is it for?

Developers adding hands-free voice control and spoken responses to an agent.

When should I use this skill?

You need wake-word voice control, speech transcription, or spoken output in an agent.

By the numbers

  • 3 STT providers: Whisper, Vosk, Google
  • built-in wake words: hey clodds, okay clodds, clodds

Files

SKILL.mdMarkdownGitHub ↗

Voice - Complete API Reference

Voice-controlled interface with wake word detection, speech-to-text, and text-to-speech.

---

Chat Commands

Voice Control

/voice start                                Start voice listening
/voice stop                                 Stop voice listening
/voice status                               Check voice status
/voice wake "hey clodds"                    Set wake word

Settings

/voice language en-US                       Set language
/voice sensitivity high                     Wake word sensitivity
/voice timeout 30                           Silence timeout (seconds)
/voice continuous on                        Enable continuous mode

---

TypeScript API Reference

Create Voice Assistant

import { createVoiceAssistant } from 'clodds/voice';

const voice = createVoiceAssistant({
  // Wake word
  wakeWord: 'hey clodds',
  wakeWordSensitivity: 0.5,  // 0-1

  // Speech-to-text
  stt: {
    provider: 'whisper',  // 'whisper' | 'vosk' | 'google'
    model: 'base',        // 'tiny' | 'base' | 'small' | 'medium' | 'large'
    language: 'en',
  },

  // Text-to-speech
  tts: {
    provider: 'elevenlabs',  // 'elevenlabs' | 'say' | 'espeak'
    voice: 'rachel',
    speed: 1.0,
  },

  // Timeouts
  silenceTimeoutMs: 3000,
  maxRecordingMs: 30000,
});

Start/Stop Listening

// Start listening for wake word
await voice.start();

// Check if listening
const isListening = voice.isListening();

// Stop listening
await voice.stop();

Event Handlers

// Wake word detected
voice.on('wake', () => {
  console.log('Wake word detected!');
});

// Speech recognized
voice.on('speech', (text: string) => {
  console.log(`User said: ${text}`);
});

// Transcription complete
voice.on('transcript', (result) => {
  console.log(`Final: ${result.text}`);
  console.log(`Confidence: ${result.confidence}`);
  console.log(`Duration: ${result.durationMs}ms`);
});

// Error handling
voice.on('error', (error) => {
  console.error('Voice error:', error);
});

// Silence detected
voice.on('silence', () => {
  console.log('User stopped speaking');
});

Speak (TTS)

// Speak text
await voice.speak('Your order has been placed');

// Speak with options
await voice.speak('Market is up 5%', {
  voice: 'josh',
  speed: 1.2,
  pitch: 1.0,
});

// Cancel speaking
voice.cancelSpeech();

Manual Transcription

// Transcribe audio buffer
const result = await voice.transcribe(audioBuffer, {
  language: 'en',
  prompt: 'Trading commands',  // Context hint
});

console.log(`Text: ${result.text}`);
console.log(`Language: ${result.language}`);

Voice Activity Detection

// Check if user is speaking
const vad = voice.getVAD();

vad.on('speechStart', () => {
  console.log('User started speaking');
});

vad.on('speechEnd', (audio) => {
  console.log('User stopped speaking');
  // audio contains the recorded speech
});

---

Wake Word Configuration

SettingValuesDescription
wakeWordAny phraseTrigger phrase
sensitivity0.0 - 1.0Detection threshold

Built-in wake words:

  • "hey clodds"
  • "okay clodds"
  • "clodds"

---

STT Providers

ProviderQualitySpeedOffline
WhisperExcellentMediumYes
VoskGoodFastYes
GoogleExcellentFastNo

---

TTS Providers

ProviderQualityVoicesAPI Key
ElevenLabsPremium30+Required
say (macOS)GoodSystemNone
espeakBasicManyNone

---

Voice Commands Examples

Once voice is active, speak naturally:

"Hey Clodds, what's my portfolio value?"
"Hey Clodds, buy 100 dollars of Trump YES"
"Hey Clodds, what are the top arbitrage opportunities?"
"Hey Clodds, set a price alert for Bitcoin at 100k"

---

Best Practices

1. Quiet environment — Background noise affects accuracy 2. Clear speech — Speak clearly for best recognition 3. Short commands — Keep voice commands concise 4. Confirmation — Enable confirmations for trades 5. Fallback — Text input always available

Related skills

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.