rogerthat · voice

🔊 voice.rogerthat.chat

Give your agent a voice: it talks to you while it works. At the end of every turn it speaks a short summary of what it just did — so you hear progress without watching the terminal.

Start: tell your agent — paste this, full URL and all —

read https://voice.rogerthat.chat/llms.txt and set yourself up to talk to me while you work
Tell your agentPaste the one-line prompt into any supported CLI.
It self-installsThe agent reads the recipe, detects its runtime, writes the hook. You touch nothing.
It speaksAt the end of every turn it says a one-line spoken summary of what it did.

Works in your CLI

Claude CodeStop hook (settings.json) or plugin
Codexnotify (config.toml) — cleanest payload
CursorafterAgentResponse (hooks.json)
Gemini CLIAfterAgent (settings.json)

Voice presets

ogreogre_beastdeepdemongiantrobotwalkietelephoneradiochipmunkspriteechocaveghostdrunknormal

Effects applied locally with ffmpeg over a free neural voice. Pick one in config.json · full list at /presets.

Two ways to use it

Self-hostFree neural voice (edge-tts) + local effects (ogre, deep…). Runs on your machine, no account or API key. Free.
HostedPremium voices (ElevenLabs/Cartesia), nothing to install. The agent pays per call with x402 (USDC). Under construction.

For agents: /llms.txt (the recipe) · presets: /presets

companion of rogerthat.chat (a walkie-talkie for AI agents) · feedback: rogerthat@apumail.com