xeon_tts

Local TTS skill using OpenVINO Qwen3-TTS for voice cloning and emotion style synthesis, supporting QQBOT workflows with strict audio length and file retentio...

This package looks functionally consistent with a local TTS/voice-clone skill, but you should not install it blindly. Before installing: 1) Inspect the xdp-tts-service package source (pip package) to verify it is trustworthy; 2) Check the model download sources — setup_env.sh sets HF_ENDPOINT to https://hf-mirror.com by default (an unexpected third‑party mirror). If you don't trust that mirror, override HF_ENDPOINT or ensure hf CLI points to official huggingface.co repos; 3) Review the full server.js (it spawns child processes and handles uploads) to see which external commands (ffprobe, etc.) it executes; 4) Be aware the installer creates user systemd services and will enable/start them and writes into ~/.openclaw/openclaw.json (a backup is created). If you want to reduce risk, run install steps in an isolated VM/container or run setup_env.sh with --skip-deps and inspect generated config files before enabling services. If anything looks unfamiliar (unknown HF mirror, unexpected pip package, or surprising exec calls), do not enable the services and seek a trusted upstream source.

Type: OpenClaw Skill Name: xeontts Version: 0.1.4 The Xeon TTS skill bundle is a legitimate implementation of a local text-to-speech and voice-cloning service using OpenVINO. The installation scripts (install.sh, setup_env.sh) automate the setup of a Python environment, download models from Hugging Face, and configure systemd user services for persistence. The Node.js gateway (server.js) handles session management, file uploads via formidable, and automated cleanup of temporary audio files. While the installation process requires significant system access (installing packages via apt/yum and downloading Miniconda), these actions are transparent and consistent with the requirements for deploying local ML inference services.