skills.rest
alamparelli
alamparelliCOMMUNITY
Explore all skills by this author →
Website𝕏@a_lamparelliGitHub

conversation

Enable bidirectional voice conversations with Claude using Push-to-Talk and VAD.

8|4|Updated Jan 11, 2026
Search Tags:#TTS#voice#conversation#STT#VAD#PTT
One-click install
npx skills add https://github.com/alamparelli/mcp-claude-say --skill conversation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: conversation
Source: https://github.com/alamparelli/mcp-claude-say/tree/main/skill/conversation
Command: npx skills add https://github.com/alamparelli/mcp-claude-say --skill conversation

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill enables natural, bidirectional voice conversations with Claude, allowing users to interact using speech for both input and output, creating a hands-free experience.

Core Features & Use Cases

  • Push-to-Talk (PTT) with VAD: Automatically starts recording when you speak and stops after a period of silence, eliminating manual key presses.
  • Full Voice Loop: Claude speaks its responses, and the system automatically listens for your next input after Claude finishes speaking.
  • Use Case: Engage in a spoken dialogue with Claude about a complex topic, asking follow-up questions and receiving spoken answers without needing to type.

Quick Start

Start a voice conversation by saying 'conversation mode'.

Frequently Asked Questions about conversation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I have a voice conversation with Claude using push-to-talk and automatic silence detection?▼

You can start a voice conversation by saying 'conversation mode'. The system uses Voice Activity Detection to automatically start recording when you speak and stops after a period of silence, eliminating manual key presses.

Can I use local STT and TTS backends for voice interaction instead of cloud services?▼

Yes, local voice processing is supported using Parakeet MLX for Speech-to-Text and Kokoro MLX for Text-to-Speech. macOS native TTS and Google Cloud TTS are also available as alternative backends for multilingual interactions.

Does the voice conversation loop automatically listen for my next question after Claude finishes speaking?▼

Yes, the full voice loop enables Claude to speak its responses, and the system automatically listens for your next input after Claude finishes speaking, providing a seamless conversational flow with minimal user intervention.

What is Voice Activity Detection and how does it work for automatic voice recording?▼

Voice Activity Detection (VAD) automatically starts recording when you speak and stops after a period of silence. This eliminates the need for manual key presses during bidirectional voice conversations with Claude.

Can I engage in multilingual spoken dialogue and receive spoken answers without typing?▼

Yes, the voice conversation system handles multilingual interactions and provides a full voice loop where you ask follow-up questions and receive spoken answers, creating a completely hands-free experience without needing to type.

Related Skills

Explore More Content & Communication Skills→
psd401psd401
2

local-tts

Generate speech audio with MLX acceleration and the Kokoro-82M model on Apple Silicon.

Official
Intermediate
reedmayhew18reedmayhew18
23

oc-voice

Transcribe voice chat messages with Whisper and route text downstream.

Community
Intermediate
refly-airefly-ai
199

fish-audio

Generate speech, transcribe audio, and create voice narration with Fish Audio models.

Official
Intermediate
anicca-labsanicca-labs

speech-recognition

Facilitate on-device voice-to-text dictation on iOS and Android using expo-speech-recognition.

Official
Intermediate