This is a wrapper around sherpa-onnx that gives you fully offline text-to-speech with no cloud dependencies. It handles the messy parts: downloads the right runtime binary for your OS, pulls down a Piper voice model (defaults to en_US lessac), and wraps the CLI so you can generate speech files locally. You'll need to set a couple environment variables pointing to where the runtime and models live, but after that it's just a simple command line call to turn text into WAV files. Useful if you're building something that needs TTS but can't rely on network calls, or if you just don't want to send your text to someone else's API.
npx -y skills add steipete/clawdis --skill sherpa-onnx-tts --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
davila7/claude-code-templates
orchestra-research/ai-research-skills
agentspace-so/runcomfy-agent-skills
inferen-sh/skills