Introduction

DeepSeek Harness (DSH) uses a plugin-based architecture. In agent development or daily use, voice input commands are more efficient than switching input methods. To address this need, the yafangwang9/dsh-voice-plugin plugin provides microphone access for the Composer component in DSH.

Plugin Overview

This is a voice input plugin for DeepSeek Harness. It allows users to use voice input through the microphone in DSH’s Composer interface and convert speech to text in real time.

  • Name: yafangwang9/dsh-voice-plugin
  • Maintainer: yafangwang9
  • License: MIT

Core Features

This plugin provides the following capabilities:

  • Voice input for DSH Composer
  • Speech-to-text conversion
  • Microphone support

Installation and Enablement

Install the plugin and enable the Web Profile.

  1. Run the installation command:
    dsh plugin --profile web add dsh-voice-plugin
  1. After installation, you must restart the dsh web process for the plugin to take effect.

Configuration and Usage

After installation and restarting, configure the model and API key in Settings.

  1. Open Settings → Plugins → voice-plugin.
  2. Select a speech recognition model:
    • For Chinese: select Doubao Voice
    • For English or other languages: select OpenAI
  3. Enter the API key for the selected model.
  4. Click the microphone icon in the Composer interface to start voice input.

If no API key is configured, clicking the microphone icon will redirect you directly to the Settings page.

Notes

  • The plugin runs with the permissions of the current DSH process. Before installing third-party plugins, it is recommended to review the source code and license (MIT).
  • This plugin is maintained by the community and has no official affiliation with DeepSeek or High-Flyer.

Summary

This plugin provides DSH users with a way to use microphone-based voice input directly in Composer. To view the full documentation or submit feedback, visit the project directory or GitHub repository.