explaingit

chendrizzy/claude-tts

Analysis updated 2026-05-18

1PythonAudience · developerComplexity · 3/5LicenseSetup · moderate

TLDR

A Claude Code plugin that speaks a filtered audio summary of your AI coding agent's activity, voicing only test results, errors, insights, and final answers while staying quiet through the noise.

Mindmap

mindmap
  root((claude-tts))
    What it does
      Speaks agent events
      Filters noise
      Summarizes long text
    TTS Engines
      edge-tts
      Kokoro local
      System say/espeak
    LLM Backends
      Ollama local
      OpenAI-compat
      No-LLM fallback
    Audience
      Developers
      Claude Code users
Click or tap to explore — scroll the page freely

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Hear test pass/fail counts spoken aloud without watching your terminal while a coding agent runs.

USE CASE 2

Get audio alerts for errors the agent encounters, so you can stay focused on other work.

USE CASE 3

Listen to the agent's final answer at the end of a task hands-free.

What is it built with?

Pythonedge-ttsKokoroOllamaClaude Code

How does it compare?

chendrizzy/claude-ttsa-bissell/unleash-liteabhiinnovates/whatsapp-hr-assistant
Stars111
LanguagePythonPythonPython
Setup difficultymoderatehardhard
Complexity3/54/53/5
Audiencedeveloperresearcherdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you get it running?

Difficulty · moderate Time to first run · 30min

Requires a running Ollama instance or another OpenAI-compatible LLM server, edge-tts is the zero-dependency TTS default.

MIT license, use, copy, modify, and distribute freely for any purpose, including commercial use.

In plain English

claude-tts is a plugin for Claude Code, the AI coding assistant from Anthropic, that reads out loud a filtered summary of what the agent is doing while it works. Instead of staying silent or printing walls of text, the tool converts key moments in the agent's activity into spoken audio you can hear without watching the screen. The core idea is selective speech. The tool listens to everything the agent does but speaks only four kinds of event: test results (like "23 passed, 4 failed"), error messages written to stderr, notable insights the agent flags, and the agent's final answer at the end of a task. Routine activity such as reading files, editing code, running git commands, or printing code blocks is ignored. A local AI model running on your own machine helps judge which moments are worth speaking and can summarize long passages before they are voiced. If you prefer not to run a local AI, the tool still works using deterministic rules alone. The text-to-speech step is handled by one of several supported engines, including a free cross-platform option called edge-tts, platform system voices on macOS and Linux, and a local neural voice model called Kokoro designed for Apple Silicon machines. The spoken output has markdown stripped out before synthesis, so you never hear punctuation symbols or formatting characters read aloud. Installation is through the Claude Code plugin marketplace. A setup command detects your operating system, picks an appropriate voice engine and AI backend, and installs a background service that starts automatically when you begin a coding session. A doctor command lets you re-check the setup at any time. The plugin requires Python 3.11 or newer and is released under the MIT license, which allows free use for any purpose.

Copy-paste prompts

Prompt 1
How do I install claude-tts as a Claude Code plugin and configure a local Ollama model as the LLM backend?
Prompt 2
What TTS engines does claude-tts support on Linux, and how do I switch from edge-tts to the system espeak voice?
Prompt 3
How does claude-tts decide which Claude Code events to speak and which to ignore?

Frequently asked questions

What is claude-tts?

A Claude Code plugin that speaks a filtered audio summary of your AI coding agent's activity, voicing only test results, errors, insights, and final answers while staying quiet through the noise.

What language is claude-tts written in?

Mainly Python. The stack also includes Python, edge-tts, Kokoro.

What license does claude-tts use?

MIT license, use, copy, modify, and distribute freely for any purpose, including commercial use.

How hard is claude-tts to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is claude-tts for?

Mainly developer.

Open on GitHub → Explain another repo

This repo across BitVibe Labs

Scan in gitsafehub Deploy in gitdeployhub chendrizzy on gitmyhub

Verify against the repo before relying on details.