v0.1.0 — Initial Public Release

VoiceClaw v0.1.0

The world's first Voice Coding Agent, powered by Gemini 3.1 Flash Live and Claude Code.

Features

• Voice-first coding — speak naturally, Gemini responds with voice
• Real-time narration — a second Gemini session provides live commentary while Claude works
• Full code agent — Claude Code reads, writes, and runs commands in your project
• Unified timeline — filterable event log showing Gemini thinking, tool calls, Claude activity, and file changes
• Git checkpoints — automatic snapshots before code changes, with voice-controlled rewind
• Session resumption — Gemini sessions persist across page reloads
• Audio visualization — wave renderer driven by real mic/speaker energy
• Screenshot support — paste or attach images for visual context
• Model switching — change Claude model and reasoning effort by voice
• Cancel task — stop Claude mid-operation by voice
• New Chat — clear all context and start fresh
• 12 voice tools: investigate, code, read, run, plan, debug, review, rewind, cancel, and more

Requirements

• Python 3.11+
• Node.js 18+
• Claude Code CLI (installed and authenticated)
• Gemini API key (free from Google AI Studio)

Getting Started

See the README for setup instructions.