heardlabs

heardlabs / heard

Public

A voice companion for AI coding agents. Speaks your agent's replies so you can keep working.

16
2
100% credibility
Found Apr 29, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Heard is a macOS application that narrates the outputs of AI coding agents aloud, allowing developers to multitask away from their screens.

How It Works

1
🔍 Discover Heard

You find Heard, a helpful voice for your AI coding buddy, on a website or from a friend.

2
📥 Get the app

Download the simple app and drag it to your Applications folder to get started.

3
🎤 Choose a voice

Pick a fun voice like a calm expert or movie butler, and connect a voice service for the best sounds.

4
🔗 Link your AI helper

Tell Heard to team up with your favorite AI coding tool so it can listen in.

5
🔊 Hear it in action

Play a quick preview to listen to the voice narrating a coding session.

6
💻 Code while moving

Start your AI coding work, walk around, and hear updates on tests, edits, and fixes.

✅ Free to roam

Now you code productively anywhere, with your AI's voice keeping you in the loop.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is heard?

Heard is a Python voice companion for AI coding agents like Claude Code and Codex, narrating tool calls, intermediate thoughts, and final summaries aloud via ElevenLabs TTS or free local Kokoro. It lets you walk away from your screen, stay in the loop on agent progress, and handle multiple sessions with distinct voices—pairing perfectly with voice input tools like Wispr Flow. Install via pipx or a one-line curl for the macOS app, then hook it with `heard install claude-code`.

Why is it gaining traction?

Distinct personas like calm Aria or witty Jarvis rewrite outputs in-character using Claude Haiku, while verbosity profiles (quiet to verbose) and auto-silencing on mic activity keep narration contextual without overwhelming. Multi-agent swarm mode prioritizes your active session, piercing failures from backgrounds with "Agent api: tests failed." The `heard demo` CLI previews voices instantly, and hotkeys (tap silence, hold replay) make it frictionless.

Who should use this?

Mac devs pair-programming with Claude Code across tabs or terminals, especially those running parallel agents on web/api repos. Ideal for voice assistant companion app fans wanting a github voice AI that handles tool-heavy workflows without babysitting outputs—think full-stackers debugging tests hands-free.

Verdict

Early alpha (v0.4, 16 stars, 1.0% credibility) but battle-tested by the author; solid docs and Apache 2.0 make it forkable. Try for Claude Code if you multitask—pin v1 for production stability. (187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.