dsh-feiyu-say: Add a "Text-to-Speech" tool to DSH

Introducing the DeepSeek Harness plugin dsh-feiyu-say, which utilizes the Alibaba Qianwen (Bailian) TTS WebSocket interface for speech synthesis. It outputs monaural WAV files and allows previewing in the settings page, remaining unbound from any Instant Messaging platform. The article details its prerequisites (Node 22, ffmpeg, Qianwen API Key), three installation commands, four core configurations, and typical usage for the three interfaces—send_voice, voice_list, and voice_status. It also covers the Qianwen interface workflow, privacy reminders regarding per-character billing, and common troubleshooting paths.