Text-to-speech for dsh web via Edge TTS
Vim0x3c/dsh-tts · skills
DeepSeek Harness (dsh web) Voiceover Plugin: Read the final results of each assistant's response in Edge TTS Microsoft Neural Voice.
★ 4GitHub stars
0Forks
2026-08-25Last updated
TypeScriptLanguage
MITLicense
Key features
- J. Edge TTSNode ends with Microsoft Edge "Loud Read" Neural Voice Interface online synthesis MP3, played by a browser - free of charge without OpenAI/third-party keys.
- Hornet switchEnter the horn button in the boxing tool line (with hints for suspension), click "Auto-read new reply" on/off, and interrupt the current reading as soon as it closes.
- Half of the host (lib/index.js)Start a small HTTP service that binds 127.0.0.1 only (detection port 8765-87880), route
- The self-fulfilled RFC 6455 WebSocket (Node Native WebSocket/ undici) filters the necessary Microsoft handshake, which results in a refusal to upgrade and hands-held return 101).
- Read-only final resultsintermediate steps, tool call instructions not read
Requirements
- Edge Cloud Synthesis requires a network: host must have access to speech.platform.bin.com.
Install command
dsh plugin --profile web add github:Vim0x3c/dsh-tts