audio-reply

Generate spoken replies on-device using local TTS from URLs or chat prompts.

Updated Jan 23, 2026

Search Tags:#audio #claude-code #tts #on-device #web-content #read-aloud #local-tts

Ready to run

npx skills add https://github.com/MaTriXy/audio-reply-skill --skill audio-reply

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: audio-reply
Source: https://github.com/MaTriXy/audio-reply-skill/tree/main
Command: npx skills add https://github.com/MaTriXy/audio-reply-skill --skill audio-reply

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill enables generating spoken audio responses entirely on-device using local TTS, removing reliance on cloud services and enabling offline playback.

Core Features & Use Cases

Read URL content aloud by fetching, extracting readable text, and converting it to speech.
Generate conversational audio responses from user prompts.
Speak any assistant reply as audio and play it on-device.

Quick Start

Install the skill in your Claude Code skills directory and trigger either a URL read or a conversational query to produce audio.

Frequently Asked Questions about audio-reply

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

FAQPage Schema

How do I generate spoken audio from web content locally without cloud services?▼

On-device TTS generation fetches URL content, extracts readable text, and converts it to speech locally. This avoids cloud dependencies and enables offline playback using local TTS models.

Can I convert assistant chat replies into audio for on-device playback?▼

Yes, conversational audio responses can be generated from user prompts. The text reply is synthesized into spoken audio and played automatically on-device using local TTS.

What dependencies do I need to run local TTS for reading text aloud?▼

Local TTS requires the MLX Audio chatterbox-turbo model and the uv CLI for generation. These enable on-device resources to synthesize and play spoken audio.

Does local TTS generation work completely offline for reading URLs?▼

Local TTS synthesis and playback work offline once the model is available. Fetching URL content still requires an internet connection to retrieve the web page before extracting readable text.

What's the best way to turn text into spoken audio without relying on cloud TTS APIs?▼

Using on-device local TTS eliminates cloud service reliance. The skill handles fetching, processing, synthesizing, playing, and cleaning up audio files entirely on-device.

Are there limitations when using on-device TTS for conversational responses?▼

On-device TTS requires sufficient local compute resources to run the MLX Audio chatterbox-turbo model. Generating conversational audio depends on having the uv CLI and model installed locally.