DeepSeek Harness PluginManifest Valid

muxiva-dsh-voice

Local-first, full-duplex voice interface for DeepSeek Harness on Apple Silicon Macs.

UI & ProductivityTerminal & TUIDeveloper ToolsBrowser & WebSecurity & Policy

2GitHub Stars0ForksUpdated2026-08-16

Source Homepage

Install

$ dsh plugin --profile web add @muxiva/dsh-voice@alpha

Plugin Overview & Capabilities

AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.

source-grounded

Muxiva Voice is a native DSH Bundle and web client plugin that bridges a local Muxiva speech pipeline with DeepSeek Harness sessions. It adds microphone voice interaction, speech recognition, TTS playback, VAD-based interruption, and a voice control UI while DSH retains agent, session, tool, and transcript responsibilities.

Key Capabilities

Adds a DSH web voice UI with microphone controls and listening, hearing, thinking, and speaking states.
Processes local PCM audio through VAD, ASR, sentence scheduling, TTS, playback, and interruption handling.
Supports Chinese and English speech recognition through SenseVoice.
Uses VAD and recognized speech to trigger barge-in, cancel prior playback, and cancel the active agent turn.
Synthesizes assistant responses with Qwen3-TTS and browser playback.
Provides setup, model download, diagnostics, headless start, and Muxiva Studio observation commands.

Useful For

Conversing with a DeepSeek Harness agent by voice on a supported Mac.
Adding local speech input and spoken agent responses to DSH web sessions.
Interrupting an in-progress spoken response by speaking again.
Diagnosing voice-pipeline behavior with Muxiva Studio observability.

Who It Fits

DeepSeek Harness web users on Apple Silicon Macs.
Developers building or evaluating local voice-agent workflows with DSH.
Users who need Chinese or English voice interaction with DSH.

Documented Limitations

Alpha release.
Requires macOS arm64, Node.js 22.19+, Python 3.11–3.13, the Muxiva 0.1.1 CLI, and the official DSH CLI.
Declared DSH peer compatibility is limited to 0.1.0-rc.5 and 0.1.0-rc.6.
Supports Chinese and English speech finals; non-speech events and some Japanese/Korean noise hallucinations are rejected.
Models are downloaded during setup and remain subject to their own licenses.

DSH Compatibility

Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.

Not tested yetNo runtime compatibility tests have been published yet.

Security Signals

Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.

Dependency Counts

package.json declares 0 runtime, 0 development, 4 peer, and 0 optional dependencies.

info

Dsh Bundle Declared

package.json declares DSH bundle metadata.

info

License Declared

GitHub reports the repository license as Apache-2.0.

info

Package Manifest Available

A root package.json was captured and can be inspected by the registry.

info

Source Available

Public GitHub source metadata is available for this registry snapshot.

info

Source & Registry Notes

Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.

Source repository: PiyotaHu/muxiva-dsh-voice
Registry source: GitHub · dsh-plugin topic
Registry classification: Plugin
Source checked: 10e5c79 · 2026-08-16

This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.