Edge TTS Chinese Voice Synthesis Service

What Problem It Solves

When OpenClaw QQBot needs to turn Chinese text into stable voice replies, directly coupling a local synthesis library or a third-party API often creates three practical issues: process crashes can take down the host bot, long text can trigger parameter or encoding failures, and timeout paths may lack recovery behavior. This skill turns Microsoft Edge TTS into a local voice service, giving QQBot a usable Chinese voice output path without replacing the bot framework.

How The Skill Works

The implementation uses a CLI-based service model: the edge-tts Python package handles synthesis, while a Node.js service exposes the voice generation interface. Bot components can call it through 127.0.0.1:5050.

• Voice selection: several Chinese Neural voices are built in, with zh-CN-XiaoxiaoNeural as the default, plus zh-CN-YunxiNeural, zh-CN-YunjianNeural, zh-CN-XiaoyiNeural, and zh-CN-YunyangNeural.
• Long-text handling: very long content is passed through UTF-8 temporary files to avoid Windows command-line length and Chinese encoding issues.
• Stability mechanisms: child-process isolation, a 60-second OS-level timeout, up to 3 retries, and exponential backoff reduce failures caused by transient network conditions.
• Configuration scope: the setup covers dependency checks, workspace deployment, openclaw.json configuration, and Windows scheduled tasks, while runtime still requires Python, Node.js 16+, and network access.

Boundaries And Caveats

This skill fits local Windows deployments where a bot needs Chinese voice replies and the 5050 port is available. It is not a good fit for offline-only environments or fully local TTS requirements. Before enabling it, confirm that edge-tts is installed, the port is free, and firewall rules allow loopback access. The skill does not claim commercial TTS quality guarantees or compliance coverage; it provides a practical synthesis path for Chinese content.