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.

CATEGORY

Vision & Media DSH plugins

Verified manifests and exact versions in this category.

All categories

All pluginsAgents & Orchestration1.6kMemory & Context2.1kCoding & Developer Tools4kUI & Customization5.6kIntegrations & Communication3kVision & Media915Search & Research1.2kSecurity & Access1.8kModels & Usage2.5kProductivity & Workflow3.6k

All plugins

915 plugins
v1.1.0
dsh-audio-visualizer

Audio Visualizer

System-audio driven UI visualizer for DeepSeek Harness: a draggable 48-band spectrum chip plus a frame-wide bass glow (memory-only FFT).

↓ 0★ 0
Updated 11/3/2026
v0.5.0
dsh-auto-sight

Auto Sight

Let a text-only model see images by automatically calling a vision model from the same provider.

↓ 0
★ 0
Updated 11/3/2026
Pv0.1.0
dsh-phonephoto-ezinput

Phonephoto Ezinput

Scan a QR code with your phone to take a photo and send the image directly to a specified conversation in DeepSeek Harness (no app installation required)

↓ 0★ 0
Updated 11/3/2026
Vv0.1.4
@lifenrir/dsh-video-preview

Video Preview

在 DSH 右侧预览面板直接播放工作区视频。Native video playback in DSH's right-side document preview.

↓ 0★ 0
Updated 11/3/2026
Sv0.1.3
@local/dsh-session-sources

Session Sources

Recover images, files and explicit links from the current conversation.

↓ 0★ 0
Updated 11/3/2026
Iv0.2.2
dsh-image-relay

Image Relay

Generate images from a text prompt through an OpenAI-compatible relay and attach them to the conversation.

↓ 0★ 0
Updated 11/3/2026
Vv1.0.0
@k-launcher/dsh-vision-control

Vision Control

DSH-Vision-Control（dsh-视-驭）: real-time visual perception + precise desktop control. computer_* control tool suite + computer_watch real-time screen perception, with one-click authorization via the composer toolbar chip.

↓ 0★ 0
Updated 11/2/2026
Sv0.1.18
shengcheng

Shengcheng

DSH plugin: generate images and videos using configured media-capable model routes.

↓ 0★ 0
Updated 11/2/2026
Sv0.2.1
dsh-super-wechat-bot

Super Wechat Bot

WeChat Chatbot · DSH Plugin: OCR screen reading + human-like timing, replying to WeChat messages in your own voice

↓ 0★ 0
Updated 11/2/2026
Vv0.2.2
dsh-video2text

Video2text

Panel converts video links into copy; tools let the agent view videos (video_text / video_frames). The two lines can be disabled separately.

↓ 0★ 0
Updated 11/2/2026
Iv1.2.0
dsh-image-size-guard

Image Size Guard

Keeps model requests under each provider's request-body ceiling by replacing the oldest history images with placeholder text.

↓ 0★ 0
Updated 11/2/2026
Ov0.1.0
dsh-openimage

Openimage

OpenAI-compatible image-generation workbench for DeepSeek Harness with built-in SenseNova U1.5 support: infinite canvas, board gallery, queue governance, mask inpaint, tiled upscale

↓ 0★ 0
Updated 11/2/2026
Lv1.11.2
dsh-longpress-shot

Longpress Shot

Long-press the left mouse button anywhere in the DSH UI to capture the window or a dragged region as PNG, saved to ~/.dsh/longpress-shots with a draggable preview card that sends the image straight into the composer.

↓ 0★ 0
Updated 11/2/2026
Dv2.3.1
dsh-dfy-mirror

Dfy Mirror

dfy蓝色大肥鱼手机镜像: Adds phone screen mirroring to DSH, using a native scrcpy window as the display, with three entry points: the DSH floating window, desktop control window, and settings page. One process, one state machine.

↓ 0★ 0
Updated 11/2/2026
Vv0.1.0
dsh-vision-capability

Vision Capability

Prove which models really read images, then declare vision support per model — with a settings panel. / 先实测模型是否真能读图，只为通过检测的模型开启图片输入，并附设置界面。

↓ 0★ 0
Updated 11/2/2026
Sv0.1.9
dsh-skin-keymark

Skin Keymark

DSH skin: static/video/generative gradient wallpapers + composition zoom and pan + full-window coverage (the skin shows through the sidebar and all panels together) + per-area opacity + text-color-only labels (writes only color, produces no boxes)

↓ 0★ 0
Updated 11/2/2026
Jv1.9.0
dsh-jinhsi-avatar

Jinhsi Avatar

A 今汐 avatar in the DSH window: the official character artwork is processed with image-to-video generation to create four sets of idle, walking, flying, and dozing animations, which are extracted frame by frame into transparent looping animations. It moves around the window on its own, changes size b

↓ 0★ 0
Updated 11/2/2026
Wv1.0.0
dsh-win-pronounce

Win Pronounce

Speak aloud in DeepSeek Harness with the Windows voices already installed on the machine: one speaker button beside the permission control, SAPI and OneCore voices switchable at runtime. / 用本机 Windows 语音朗读：对话输入栏权限控件右边一枚喇叭键，SAPI 与 OneCore 音源随时切换。

↓ 0★ 0
Updated 11/2/2026
Dv0.3.0
dsh-docx-from-images

Docx From Images

Screenshot -> Word typesetting for DeepSeek Harness: docx_typeset / docx_crop tools plus the image-to-docx skill.

↓ 0★ 0
Updated 11/2/2026
Av0.7.0
dsh-auto-convert

Auto Convert

Prepare files for reading images and documents: downsample oversized images, convert BMP/TIFF to PNG, and convert docx/xlsx/pptx/pdf/csv/html in place to Markdown, reaching a conclusion in one conversation.

↓ 0★ 0
Updated 11/2/2026
Vv1.6.2
dsh-voice-unified

Voice Unified

DSH unified voice plugin: multi-voice cloning and evolution, on-demand voice replies, seven-emotion TTS, direct output for cover song voice conversion, music cache integration, and management interface.

↓ 0★ 0
Updated 11/2/2026
Rv0.1.0
@lucagiftzek/dsh-request-image-cap

Request Image Cap

Clamp the per-side pixel size of every image DSH projects into a model request (Anthropic refuses >2000 px per side once a request carries many images).

↓ 0
Updated 11/2/2026
Vv2.0.0
@lucagiftzek/dsh-voice

Voice

Read-aloud, dictation and language-aware speech for the DeepSeek Harness web UI: many TTS/STT providers at once (EU vendors first: Mistral, Speechmatics, Gladia, ReadSpeaker; plus ElevenLabs, Azure, Google, Amazon, OpenAI and OpenAI-compatible servers, Deepgram, Cartesia, AssemblyAI, Fish, self-host

↓ 0★ 0
Updated 11/2/2026
Rv0.1.0
dsh-request-image-cap

Request Image Capalt · lucagiftzek

Clamp the per-side pixel size of every image DSH projects into a model request (Anthropic refuses >2000 px per side once a request carries many images).

↓ 0
Updated 11/1/2026
Pv0.1.0-beta.10
dsh-pdf-reader

Pdf Reader

AI-powered PDF reader for DeepSeek Harness with annotations, multi-PDF workflows, mixed image-text evidence, and on-demand OCR.

↓ 0★ 0
Updated 11/1/2026
Iv0.2.1
dsh-image-insight

Image Insight

Structured image recognition plugin combining local sharp-based perception with cloud vision models: one call returns descriptions, OCR, object localization, and fine-grained attributes, with image fingerprint caching for millisecond responses on repeated images.

↓ 0★ 0
Updated 11/1/2026
v0.0.4
dsh-voice-typing

Voice Typing

Voice input and conversation read-aloud: tap or use a keyboard shortcut to switch to voice mode, with text appearing as you speak; offline local recognition (SenseVoice + Silero VAD), support for custom vocabularies, hotword hints, and multiple model tiers; replies can be read aloud using three engi

↓ 0★ 0
Updated 11/1/2026
Av0.1.0
dsh-audio-tools

Audio Tools

DSH audio tools: probe, convert/extract audio tracks, and transcribe speech with your own ffmpeg + whisper.cpp or any OpenAI-compatible transcription API.

↓ 0★ 0
Updated 11/1/2026
Vv0.4.0
dsh-voice-v

Voice V

Read-aloud speaker button under every assistant message in the DeepSeek Harness web GUI, voiced by edge-tts: seven Russian and multilingual voices, the Yandex SpeechKit voices, the narration preset picked from the tone of the message, and a spectrum inside the button while the voice plays. Ships a c

↓ 0★ 0
Updated 11/1/2026
v1.4.0
dsh-copilot-vision

Copilot Vision

A local Windows Computer Use enhancement plugin for DSH (DeepSeek Harness): pure-JS visual differencing, grid position encoding, keyboard injection, semantic memory compression, long-horizon task loops, and human-machine co-pilot authorization.

↓ 0★ 0
Updated 11/1/2026
← Previous1…1415161718…31Next →
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in