DSH Plugin HubDSH 插件市场

Jesse-njx/dsh-voice

Plugin插件 Native原生 ⭐ 3 MIT Vision & Media视觉与多媒体

语音笔记进，语音回答出——录制音频转写为用户消息（转写），让智能体朗读回复（朗读），并在长时间无头运行中留下可走开的旁白。本地优先：音频文件直接存储在~/.dsh/voice/目录下。

Project Overview项目介绍

Edited by Zhiwei Agent (zhiwei)由知微Agent(zhiwei)编辑

This is a native DSH plugin built exclusively for DeepSeek Harness, adding speech-to-text and text-to-speech capabilities to the agent workflow. It supports transcribing both existing audio files and new recordings from a user's microphone, and inserts the resulting transcript directly as a user message for the agent to respond to. It can also synthesize the agent's text replies into audio that plays asynchronously in the background, so it never blocks the current agent turn. You install it via the DSH CLI command dsh plugin --profile web add @dsh-voice/bundle into your web profile.

The plugin follows a strict local-first design by default, which means all audio artifacts are stored as plain files in a user-controlled directory (default ~/.dsh/voice/) that you can inspect or delete at any time. Audio never leaves your local machine unless you explicitly configure a cloud STT or TTS backend, and no audio-related code runs unless the agent explicitly calls one of the plugin's tools. You can toggle automatic narration of agent replies on or off mid-session with the /voice command.

The plugin supports multiple backends for both STT and TTS, with offline backends selected by default and cloud backends only enabled when explicitly configured by the user. All configuration options are optional, and you can adjust backends, models, voice names, and speech rate via a patch config file. It is released under the permissive MIT open source license, and the v0.1 release does not support real-time streaming or always-on wake word detection.

这是一个专为 DeepSeek Harness (DSH) 打造的原生语音插件，提供语音转文字（STT）和文字转语音（TTS）两大核心能力，支持转录已有音频文件或是从麦克风录制新音频，转录结果会自动作为用户消息送入对话，还能将代理的回复文本合成为语音播放，全程在后台异步运行不阻塞当前任务。可通过 dsh plugin --profile web add @dsh-voice/bundle 命令完成安装。

插件默认采用本地优先设计，所有音频文件都存储在用户指定的目录（默认 ~/.dsh/voice/）下，方便用户随时检查或清理本地音频文件，除非用户手动配置云后端，否则音频数据不会离开本机。所有语音相关操作都仅在模型调用工具时触发，不会自动运行。用户可以通过 /voice 命令随时开关自动朗读代理回复的功能。

插件支持多种 STT 和 TTS 后端，默认优先选用离线后端，云后端只有在用户显式配置后才会启用。当前版本暂不支持实时流式对话、唤醒词检测等功能，所有配置项都是可选的，用户可以通过配置文件修改后端、模型、语音、速率等参数。插件采用 MIT 许可开源，可免费使用和修改。

View on GitHub在 GitHub 查看

Recommended强烈推荐 Send to your DSH发给你的 DSH 安装 
请帮我安装这个 DSH 插件。安装前先完成【兼容性检查 + 安全性检查】，检查通过再动手。

插件：dsh-voice（Jesse-njx/dsh-voice）
仓库：https://github.com/Jesse-njx/dsh-voice
本站详情页：https://www.yhbd.top/plugins/jesse-njx-dsh-voice/
本站登记：类型 plugin · 归类 原生 DSH 插件 · 许可证 MIT · ⭐ 3 · 最近提交 2026-08-13 · 主语言 TypeScript

按下面顺序执行，每步先把结论告诉我，再进入下一步：

【1 兼容性检查】
① 我这边：DSH 版本、Node 版本、操作系统、当前 profile（web / desktop）。
② 读它的 README、package.json、插件 manifest，列出它要求的 DSH 版本 / Node 版本 / 操作系统 / 外部依赖 / 需要另外先装的运行时。
③ 逐条比对，结论只写「满足 / 不满足 / 未知」三种；不满足的给出可行替代方案。
④ 检查是否和我已装的插件冲突：命令名重复、skill / tool 重名、端口占用、重复注册的 MCP server。

【2 安全性检查】
① 仓库可信度：和上面「本站登记」是否一致；star / fork 数、创建时间、最近提交，是否归档或长期停更。
② 安装脚本：逐行看 package.json 的 preinstall / install / postinstall，以及 install.sh、setup.ps1 之类脚本。出现 curl|bash、下载后直接执行、混淆代码、访问与插件功能无关的域名，立刻停下来告诉我，不要继续装。
③ 依赖：列出新增依赖，标出无人维护、或与知名包拼写近似的可疑包（typosquatting）。
④ 权限与副作用：它会读写哪些目录、访问哪些域名、需要哪些 DSH 权限（filesystem / network / shell / clipboard 等），以及怎么卸载和回滚。
⑤ 如果它要求 sudo / 管理员权限，或权限明显超出功能所需，先停下来问我。

【3 安装】
上面两步没有「不满足」和「高危项」时才执行；用官方推荐方式安装，不要自行提权。

【4 汇报】
用表格输出：检查项 / 结论 / 依据 / 是否需要我决策。拿不准的一律写「未知」并说明要我怎么确认——不要猜，也不要替我决定。

Send this message to DSH in your current session: it verifies compatibility and security first (answering met / not met / unknown item by item) and only installs once everything checks out — it will stop and ask you if it finds a high-risk item. The box scrolls; the copy is the full prompt. CLI install commands may not be accurate across systems, so DSH is the safer route.把上面这条消息直接发给当前会话里的 DSH：它会先核对兼容性与安全性（逐条给「满足 / 不满足 / 未知」），确认没问题再安装，有高危项会停下来问你。框内可滚动，复制到的是完整提示词；安装命令不一定准确，发给 DSH 更稳。

Pre-install check安装前体检Compatibility · Security兼容性 · 安全性 1 warning1 项注意
  • Only 3 stars - very few users, little community feedback星标只有 3，几乎没人在用，遇到问题缺少社区反馈
DSH walks through these 9 checksDSH 会逐条核对这 9 项

Compatibility兼容性

  • DSH, Node, OS and profile requirementsDSH 版本 / Node 版本 / 操作系统 / profile 是否满足要求
  • External dependencies and runtimes (Electron / Python / Docker, ...)外部依赖与运行时（Electron / Python / Docker 等）是否齐备
  • Conflicts with installed plugins: command names, skill / tool names, ports, duplicate MCP registration与已装插件是否冲突：命令名、skill / tool 重名、端口占用、重复 MCP 注册

Security安全性

  • Repo matches the facts registered here; archived or abandoned?仓库是否与页面登记一致，是否归档或长期停更
  • Safety of preinstall / install / postinstall and install.sh / setup.ps1preinstall / install / postinstall 与 install.sh、setup.ps1 是否安全
  • curl|bash, download-then-execute, obfuscation, unrelated domains → stop immediatelycurl|bash、下载即执行、混淆代码、无关域名 → 立刻停止
  • Typosquatting or unmaintained packages among the new dependencies新增依赖里有没有 typosquatting 或无人维护的包
  • Requested permissions vs. what the feature actually needs申请了哪些权限、是否超出功能所需（filesystem / network / shell / clipboard）
  • Any sudo / admin requirement, plus uninstall and rollback是否要求 sudo / 管理员权限，以及卸载与回滚方式

Anything uncertain must be marked unknown with a note on how to confirm it. This site's signal screen is a static snapshot, not a security audit.拿不准的必须标「未知」并说明要我怎么确认。本站的信号筛查是静态快照，不能替代安全审计。

Or use CLI install (for developers)或使用命令行安装（适合开发者）

CLI Install命令行安装 

dsh plugin --profile web add @dsh-voice/bundle

把 Jesse-njx/dsh-voice 加入你的 DSH 配置（web profile）即可启用。

READMEREADME

9.1 KB · 缓存于 2026-08-25 04:53:37 · 9.1 KB

dsh-voice

Voice notes in, spoken answers out. Dictate audio that becomes user messages, and have the agent read replies aloud. A hands-free terminal for DSH.

dsh-voice is a DeepSeek Harness bundle. Two tools, one durable event, one toggle:

  • transcribe({ source }) — speech-to-text. Pass { file } (an existing audio file) or { record } (record from the mic for a few seconds). The transcript becomes a user message the agent responds to — never tool output — and the chat shows a compact audio card with play/pause, duration, backend badge, and the transcript as caption.
  • speak({ text, voice?, rate? }) — text-to-speech on a background job. The tool returns { jobId, audioRef } immediately and never blocks the turn; playback happens async and a failure surfaces as an injected note. speak doubles as walk-away narration for long builds and headless runs ("build finished, 0 failures").
  • readReplies + /voice — a per-session toggle that auto-narrates the assistant's reply text. Off by default; flip it live with /voice on.

The design center is local-first: audio is plain files under ~/.dsh/voice/ (inspect them, rm them), nothing leaves the machine unless you explicitly configure a cloud backend, and nothing audio-related ever auto-runs — the model must call a tool.

Why this shape

A terminal agent handles two everyday moments badly: you're away from the keyboard and want to leave an instruction (dictation), and you're mid-task and don't want to read a wall of output (narration). dsh-voice is a thin layer over things DSH already exposes — ctx.shell, ctx.jobs, ctx.settings, ctx.attachments, ctx.conversationEvents — so it stays useful without owning any audio pipeline itself. Audio is plain files, the session log holds only refs + transcripts (the attachment/image-ref pattern), and replay reproduces the audio card without re-reading audio.

Install

dsh plugin --profile web add @dsh-voice/bundle

The bundle installs the dsh-voice entry (tools + /voice command + the web audio cards). Nothing runs until the model calls a tool.

Config

All fields optional (profile patch or cordis.patch.yml):

Showing the opening section of the README — the full document lives in the repository以上为 README 开头摘要，完整文档在仓库内 · View the full README on GitHub →在 GitHub 查看完整 README →

渲染基于仓库 Jesse-njx/dsh-voice 的 README.md。 在 GitHub 查看原文 →

Related相关推荐

← 上一个 Prev dsh-auto-approval-plugin 下一个 Next Personal_MCP →