微信聊天分析及预测工具：本地聊天记录分析与语音交互
SkillHub 技能 @user_9d5a2a39/wechatanalyzer 在本地解析微信等聊天记录，提供 MBTI、大五人格、情感趋势、关系健康度、风险与隐私扫描、群聊与跨会话对比、多层对话预测及 Markdown/CSV/HTML 导出；v2.9.0 起支持 voice 子命令与 Windows SAPI 语音降级。分类 data-analysis，无需 API Key 即可使用核心能力。安装参考 skillhub.cn/install/skillhub.md，CLI 依赖 pip 与 analyze-v2、session、capture、predict-v2 等命令。分析结果为启发式估计，敏感场景需人工复核。
Read Moredsh-live2d-voice: A Live2D Voice Companion for DeepSeek Harness
`dsh-live2d-voice` is a DeepSeek Harness (DSH) plugin designed to enable real-time voice interaction and dynamic visual expression through Live2D models. Built on Cubism 4 and pixi.js rendering, it supports streaming speech synthesis, continuous voice input, emotion expression synchronization, and bilingual subtitles. Users can install and configure it in the DSH interface, converse via the Live2D tab, and switch character voices using the voice library. Features include immersive fullscreen mode, gaze tracking, and per-workspace override configuration. When using it, be aware of the session-level restrictions on prompt injection and its network dependency.
Read More@xp1024/dsh-voice-gateway: Voice Interaction Plugin for DSH Web Client
`@xp1024/dsh-voice-gateway` is a plugin for the DeepSeek Harness (DSH) Web client that enables physical voice interaction capabilities. It features core functions such as six-state real-time status indication, voiceprint filtering management, two-way voice interaction, dynamic gateway configuration, and button mounting. Users need to start the local voice gateway daemon first, then install and activate the plugin via a command, after which they can access convenient voice control and status feedback in the Web interface.
Read MoreDeepSeek Harness Plugin: Retrieving Execution Results via an ElevenLabs Voice Callback
This plugin integrates DeepSeek Harness (DSH) with the ElevenLabs voice agent, addressing the challenge of inefficient log review and approval command handling on remote or mobile devices. Users can receive a call link via terminal, QR code, or push notification, then use voice on their mobile phone to listen to execution summaries, answer questions, and issue approval commands. The plugin supports voice interaction, result summarization, and command return, requiring Node.js 22, a specific DSH version, and an HTTPS environment.
Read Moredsh-voice-mimo: Voice Interaction and Tool Plugins Based on MiMo
The `dsh-voice-mimo` plugin is built on Xiaomi MiMo technology and provides a complete voice interaction solution for DeepSeek Harness (DSH). It integrates browser-based voice UI (microphone input and TTS playback) along with Agent tools (speech transcription, understanding, and synthesis), enabling users to interact with the system using voice. The plugin is now archived, and its core functionality has been migrated to the `dsh-mimo-agent-tools` repository. Developers need to configure `XIAOMI_API_KEY` and refer to the `./client` and `./mimo` directories in the new repository for the latest documentation.
Read MoreDeepSeek Harness Live Voice Plugin Overview
`jstn-1g/dsh-live-voice` is a plugin for DeepSeek Harness (DSH) designed to provide a controlled voice interaction layer. It features session isolation, explicit authorization, single-voice-turn control, and draft handoff. The plugin allows users to manually trigger interactions and place transcribed text into a draft area, but it does not automatically submit content to the editor or execute tools. Credential-based models and physical microphone functionality, where applicable, have not yet been verified, and concurrent write limits should be considered.
Read Moredsh-kitt-voice: Voice Interaction Plugin for DeepSeek Harness
`dsh-kitt-voice` is a client-side voice interaction plugin designed specifically for DeepSeek Harness. It supports global hotkeys, multi-page routing, and offline speech synthesis, with features including real-time speech detection, barge-in responses, and multilingual neural voices. The plugin uses a Push-to-Talk mechanism to provide an immersive conversational experience, and supports independent device selection and speech rate adjustment. After installation, a full restart of Harness is required; currently, response text is sent to the cloud for processing.
Read Moredsh-voice-talk: Voice Conversation Plugin for DeepSeek Harness
`dsh-voice-talk` is the voice plugin for DeepSeek Harness, designed for hands-free use. Through full-screen voice interaction, it enables simultaneous speaking and listening, as well as real-time content generation and read-aloud. It supports Chinese and English, streaming interaction, and adjustable speech rate and voice timbre. Before using it, install host version 0.1.0-rc.7 or later and configure an Alibaba Cloud Bailian API Key to enable speech recognition and synthesis. Plugin keys are stored locally and follow the principle of least privilege, making it ideal for efficient use in Chrome or Edge browsers.
Read MoreDeepSeek Harness Voice Call Assistant Plugin
`dsh-voice-call` is a personal voice call assistant plugin for DeepSeek Harness. It enables conversation continuity across restarts through a dedicated workspace, persistent sessions, and a floating ball panel. The plugin integrates FunASR for speech recognition, supports multiple text-to-speech engines such as MiniMax and MiMo-V2.5-TTS, and provides MiMo Director Mode and voice cloning. In addition, the main session can use the `voice_task` tool to dispatch tasks to independent sessions for execution and automatically announce the results. It can be installed with a single command, and pairing it with a local ASR is recommended to reduce latency and protect privacy.
Read More@allmodels/dsh-speech: Speech Plugin for DeepSeek Harness
DeepSeek Harness introduces the `@allmodels/dsh-speech` plugin, leveraging a plugin-based architecture to provide developers with voice interaction capabilities. The plugin supports real-time speech-to-text input, spoken answer summaries, and brief speech output, requiring a connection to the AllModels.io service for authentication. Developers can install and enable it with a simple command and configure microphone and recognition parameters in the settings. It is suitable for scenarios that require quick interaction or listening to agent work summaries, simplifying the voice interaction workflow.
Read Moredsh-chatvoice: Adding voice input and audio responses to dsh web
dsh-chatvoice is a DSH plugin maintained by FuzzySoul, designed to optimize the interaction experience of the DeepSeek Harness (dSH) web interface and realize a "voice input + AI reply reading" closed loop. Based on the browser's native Web Speech API, the plugin features zero configuration, zero cost, and no API key requirement, with no network requests or sub-processes. Core functions include voice input via the microphone, with recognition results appended to the input box in real-time and support for mid-conversation typing modifications. It supports one-click or automatic reading of the assistant's replies, which can be interrupted at any time. A settings page is provided to configure recognition language, voice, and speed; for Chinese scenarios, it defaults to the Edge built-in natural voice. The plugin features a comprehensive error notification mechanism that provides friendly prompts for situations such as permission denials or browser incompatibility.
Read Moredsh-voice: Voice input to user message, readable replies.
The DSH plugin `dsh-voice` aims to enhance terminal agent efficiency through voice interaction, implementing speech-to-text, text-to-speech, and reading aloud. Adhering to a "local-first" philosophy, the plugin stores audio files locally by default and does not automatically record or upload to the cloud unless users explicitly configure cloud backends such as OpenAI or Edge-TTS. Its core functions include: `transcribe`, which supports transcription from files or real-time microphone recording, inserting results into the session as voice cards; `speak`, which synthesizes and plays text in the background, suitable for announcing long task status; and the `/voice` command, used to toggle session-level auto-read replies (defaulting to off). The plugin is compatible with Node.js >=20, supports local backends like Whisper-local and macOS Say, and can work with `dsh-cr`.
Read More