# dsh-voice-plugin

[English](README.md) | [中文](README.zh.md)

> If you are an agent, read AGENTS.md first.

Add a microphone to the DeepSeek Harness composer so you can input by voice.

## Install

```sh
dsh plugin --profile web add dsh-voice-plugin
# or straight from GitHub:
dsh plugin --profile web add github:yafangwang9/dsh-voice-plugin
```

Then restart `dsh web`.

## Configure

Open **Settings → Plugins → voice-plugin** and pick your model:

- Chinese → **Doubao Voice**
- English or other languages → **OpenAI**

Fill in only that model's API Key. If it isn't configured yet, clicking the mic just points you to Settings.

## License

MIT