AI Agent Hub
Back to plugins
dsh-dictate preview

dsh-dictate

Model Inference Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install franksong2702/dsh-dictate

Paste the following prompt into your AI chat to install this plugin:

在 DeepSeek Harness 中使用命令 dsh plugin install franksong2702/dsh-dictate 安装此插件,源码地址为 https://github.com/franksong2702/dsh-dictate。

About this plugin

Voice input often feels clunky when using text-based AI tools. dsh-dictate solves this by seamlessly integrating speech-to-text into DeepSeek Harness Composer, allowing users to capture ideas effortlessly without the friction of switching between applications.

Its core strengths lie in "context-aware" processing and "user control." The plugin automatically extracts visible text from the current session to build a temporary vocabulary, significantly improving accuracy in mixed-language transcription. Whether using the default Web Speech API or a one-click installed local SenseVoice endpoint, it ensures users have full control to edit, polish, and send results on their own terms, avoiding the passive interruptions of real-time voice bots.

This plugin is ideal for creators, developers, and professionals who interact frequently with AI. Whether you are taking meeting notes, drafting technical documentation, or simply looking to speed up your workflow, it provides a smooth and accurate voice input experience while keeping you firmly in control of the final output.

Screenshots

Use Cases

  • Meeting minutes
  • Writing assistance
  • Fast input

Best For

  • Creators needing efficient input
  • Privacy-conscious developers
  • Users needing text polishing