DeepSeek Harness Plugin Hub

Publish and manage complete Harness Profiles. Discover Plugins for your next setup.

Explore

PluginsPresetsDocsNews

Community

Publish a pluginContactReport an issue

Resources

Plugin Hub on GitHubDeepSeek HarnessSystem statusPrivacy notice
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

Independent and unofficial. Not affiliated with, authorized by, or endorsed by DeepSeek.

Voice Context — DSH Plugin for DeepSeek Harness
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in
← Plugins
V

dsh-voice-context

Voice Context

Voice-to-text for DeepSeek Harness: a composer mic button backed by a cloud STT provider (SiliconFlow) or a bundled offline backend (FunASR / faster-whisper).

The plugin will be installed here. Keep web if you are unsure.

npx -y @deepseek-ai/dsh plugin --profile web add github:CharlesLiuZC/dsh-voice-context#b13083dd4f7ca72610971fd668622273f5e6a196
READMECompatibilityVersions

Compatibility and provenance

Voice Context is published as dsh-voice-context and currently resolves to version 1.0.0. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
web
Release source
github
Registry updated
10/2/2026

Versions

1.0.0stable
10/2/2026

Related plugins

Loading related plugins…

Latest
1.0.0
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
Unavailable
Files
Unavailable
Surface
web
License
MIT
Source
github
GitHub
★ 0
Weekly downloads
0
Last push
10/2/2026
View source ↗
README badge

Click the badge to copy Markdown for your README.

Do you maintain this Plugin?Claim benefit · Priority security scan

Verify the GitHub repository declared in package.json to manage this listing. After you claim it, Hub will prioritize a security scan of the current version and publish the result when it passes.

Claim this Plugin →
Report an issue

Related plugins

More verified plugins in vision-media.

Vision Routerdsh-vision-routerEyes for text-only DeepSeek Harness agents: built-in free vision chain + pixel-level tools. Remote vision providers, including the default free fallback, receive image content unless strict local-only use is configured.Imagegen@dickpy/dsh-imagegenAI image generation plugin for the dsh web GUI: text-to-image and image-to-image through configurable provider channels (gpt-image-2 / grok-imagine-image / nanobanana series / seedream-5.0-pro / dall-e-3, with native xAI Grok Imagine, Google Nano Banana aExperimental Voice Input Bundle@deepseek-ai/dsh-experimental-voice-input-bundleExperimental voice input with local SenseVoice; downloads its runtime on first useCodexdsh-codexChatGPT OAuth, Codex models, search, read_image URL support, and configurable GPT Image generation for DeepSeek Harness

README

dsh-voice-context

DeepSeek Harness(桌面端 / Web)语音转文字插件 —— 在输入框左侧加一个麦克风按钮,说话即转写进草稿;支持云端 API(SiliconFlow SenseVoiceSmall)与本地离线(FunASR / faster-whisper)两条路径。

  • 符合 DSH 插件规范:dsh.bundle(宿主补丁层)+ dsh.client(浏览器半),一个包双面。
  • 通信走 Connection RPC 通道(/voice-context),与官方 Remote 同一条受信任传输,Web GUI 与 Electron 桌面端通用,无需改动任何一方包。
  • 可直接从 GitHub 安装。

安装

桌面端自带 dsh 命令:<安装目录>\resources\runtime\cli\bin\dsh.cmd

# 从 GitHub 安装(推荐)
dsh plugin --profile desktop add github:CharlesLiuZC/dsh-voice-context

# 或从本地目录安装
dsh plugin --profile desktop add /path/to/dsh-voice-context

安装后重启应用(宿主在启动时组装插件层),输入框左侧即出现麦克风按钮。

卸载:

dsh plugin --profile desktop remove dsh-voice-context

使用

  1. 点输入框左侧的 麦克风 → 说话 → 再点一次停止;转写文本自动追加到当前草稿。
  2. 在 设置 → 语音输入 里选择处理方式:
    • 云端 API:填 SILICONFLOW_API_KEY(写入 DSH 凭据域,不落响应、不回显)。
    • 本地离线:无需密钥,走本机 127.0.0.1:8000 的后端。

首次使用建议先选一种方式,麦克风会按该选择逐次转写。

配置

在 profile 的 cordis.patch.yml 里覆盖 voice-context 行(默认值见下):

- id: voice-context
  name: 'dsh-voice-context'
  config:
    baseUrl: https://api.siliconflow.cn   # OpenAI 兼容 STT 服务地址
    model: FunAudioLLM/SenseVoiceSmall    # 云端默认模型
    language: zh                          # 语言提示
    localPort: 8000                       # 本地后端端口
    pythonBin: python                     # 本地后端使用的 Python
    apiKeyEnv: SILICONFLOW_API_KEY        # 凭据引用名
    maxBytes: 26214400                    # 音频上限(字节)
    timeoutMs: 60000                      # 上游超时

本地离线后端

本包内置一个 OpenAI 兼容的本地 STT 服务(local/funasr/),由 /voice-local 命令管理:

/voice-local status    # 硬件与依赖自检
/voice-local install   # 装 funasr + faster-whisper + CPU torch
/voice-local start     # 在 127.0.0.1:8000 启动
/voice-local stop

模型选择(兼顾速度与准确度):

引擎中文速度体积说明
FunASR SenseVoiceSmall(默认)高极快(非自回归)~1GB中文首选,CPU 可跑
faster-whisper small中快~1GB多语言、速度优先
faster-whisper medium中高中~3GB多语言、准确优先
faster-whisper large-v3高慢~6GB多语言、最高准确

local/funasr/ 也可独立部署(不经 DSH):见该目录的 README,暴露标准 POST /v1/audio/transcriptions。

工作原理

浏览器半 (lib/client.js)               宿主半 (lib/index.js)
  ├─ conversation.input.left 麦克风      ├─ connection.rpc.handle('/voice-context', …)
  ├─ settings.section 设置页      ──►    │    └─ transcribe 端点
  └─ connection.rpc.call('/voice-context','transcribe',{args})
                                          ├─ 凭据域解析 API Key(每次请求)
                                          ├─ 云端:转发 /v1/audio/transcriptions
                                          └─ 本地:转发 127.0.0.1:<localPort>
  • 通道 authority: 'loopback':仅本机可用,密钥与本地后端不会暴露给其他来源。
  • 音频以 base64 走 JSON(浏览器录 WAV,16kHz 单声道);上限由 maxBytes 约束。
  • 转写是人类输入,不进入模型的 tool 调用。

开发

pnpm install
pnpm build      # 产出 lib/index.js(宿主,ESM)+ lib/client.js(浏览器 bundle)
pnpm typecheck

lib/ 是构建产物;本仓库已提交构建结果,因此从 GitHub 安装无需在本地再构建。

许可

MIT