DSH Marketplace
中文

dsh-pet

PC2005-cloud/dsh-pet#dsh-pet

Desktop pet for the DSH Web UI with 25 transparent animations, screen wandering, click reactions and drag, plus a reproducible asset-generation pipeline.

885JavaScriptMITSource

Install

Add dsh-pet to DeepSeek Harness

This plugin lives in a subdirectory of a larger repository, and `dsh plugin add` cannot reach it — it forwards to pnpm, which reads everything after `#` as a git branch or commit. Clone the repository and add the plugin through DSH's plugin panel using a repository source, or ask the author to publish it to npm.

Due diligence

Before you install dsh-pet

  • Source of record: PC2005-cloud/dsh-pet — present in the community registry that DSH's own plugin market installs from.
  • Licensed under MIT.
  • A listing here is not a security review. Plugins run with your agent's permissions.

The AI take

What it is — The plugin provides a desktop pet that runs in the DeepSeek Harness Web UI, with transparent animations, screen wandering, click reactions, and drag support.

Who it is for — Users who run the DeepSeek Harness Web interface and want a desktop pet that wanders on screen and reacts to clicks should install this plugin. Those who do not need screen movement or click responses should not install it.

Watch out — No obvious issues were found. Sandbox testing has not been conducted yet. When installing from source code, the build script requires manual approval first.

The verdict — I would install it because it includes a reproducible asset generation pipeline that allows creation of custom animations.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.

What dsh-pet does

Translucent animation tiles flow through a selector and double-buffered frames into a wandering silhouette.

dsh-pet is a DeepSeek Harness plugin that adds an animated desktop pet to the DSH Web UI. It provides 51 transparent video animations, including idle breathing, movement, turning and reactions, and places the pet in the interface rather than adding an agent tool or business feature.

The plugin has a host-side route, lib/index.js, which exposes /pet video resources, and a browser-side client in lib/client.js. The client selects animations in sequence, mirrors them for left- or right-facing movement, and uses double-buffered playback for transitions without a blank frame. Its packaged playback assets are the 640×360 WebM files in assets/thumb/. The pet can wander within the screen, respond to clicks, and be dragged to another position. The documented selection probabilities are 30% idle, 10% turning, 40% action and 20% movement.

It is intended for developers who want a non-functional visual companion in the DSH Web UI. It does not query an LLM or API, monitor agent state, or provide weather or system information, so it is the wrong choice when those integrations are required. It is also a Web UI plugin, not a general desktop application. Recreating or customising the assets requires Python 3, ffmpeg, numpy and scipy, plus source videos or the repository’s released assets-videos files.

dsh-pet documentation

How it works

The plugin is split between a host component and a browser component:

  • lib/index.js provides the /pet video route.
  • lib/client.js runs the animation chain in the browser.
  • assets/thumb/ contains the packaged 640×360 WebM playback variants.

The client chooses another animation when the current one finishes. It supports idle, turning, action and movement states, with documented probabilities of 30%, 10%, 40% and 20% respectively. Animations can be mirrored to face either direction. Double buffering cross-fades between clips, while shared foot alignment keeps the pet positioned on the floor line.

Interaction and movement

The pet appears in the DSH Web UI and starts at the lower-right area. It can wander across the screen, checks available space before moving, and avoids leaving the screen. Clicking triggers a response animation, and dragging allows the pet to be repositioned.

The README describes this as a visual-only plugin. It does not add weather lookup, system monitoring or agent-state awareness, and runtime does not require an LLM or API call.

Asset-generation pipeline

The repository includes a reproducible pipeline for creating replacement assets. Place 51 source .mp4 files in video/, then run these commands from scripts/:

python watermark_step01.py
python chroma_step02.py
python normalize_step03.py
python encode_thumbs.py

The stages produce intermediate step01/ through step04/ output. The process masks watermark areas, removes the green screen using HSV hue processing, normalises clips to 2160×1215, and creates 640×360 playback variants. Copy generated .webm files into dsh-pet/assets/thumb/ to use them in the plugin package.

Requirements and scope

The asset scripts require Python 3, ffmpeg, numpy and scipy; the scripts use workspace .tools/ for ffmpeg according to the README. Source videos are not stored in Git and can be obtained from the assets-videos GitHub release. The documented runtime target is the DeepSeek Harness Web UI. No plugin configuration keys or environment variables are documented.

Written from the project's own documentation and kept in sync with it. Where the two disagree, the source is authoritative — read the README on GitHub

Same category

Alternatives to dsh-pet

dsh-ads

Nagi-ovo

481

Parody ads in 2005-Chinese-web style: sidebar banners, in-chat feeds, corner popups, and a close button whose hit area is smaller than it looks. All fictional.

Installed cleanly when we ran it

GitHub sourceTypeScriptyesterday

AI review

dsh-ads

What it is — Inserts 2005 Chinese web style fictional ads into DeepSeek Harness.

Who it is for — If you want to add fun 2005 style ads to sidebar and chat in web profile, install it. Users who prioritize inference smoothness don't need to install it.

Watch out — Sandbox test passes when installed to a new profile and registers successfully. No obvious issues found.

The verdict — I would install it because it allows closing individual ad slots in settings, but only if you accept temporary pauses to inference.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
Details

whale-girl

vlln

225

Desktop pet (QQ-pet style): floats in the corner, draggable, feedable, playable.

Installed cleanly when we ran it

GitHub sourceJavaScriptyesterday

AI review

whale-girl

What it is — DSH Web GUI right-bottom floating desktop pet, draggable, feedable, playable and accumulates task experience into ranks.

Who it is for — Users doing tasks in DSH Web GUI are suitable to install it to accumulate ranks. Users not wanting to add desktop companions can skip it.

Watch out — No obvious pitfalls found. Sandbox test passed in new profile. Configuration changes like size and walk are hot-reload without restart.

The verdict — I would install it because it accumulates ranks from tasks but only for DSH Web GUI users.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
Details

dsh-dafeiyu

QCYTSN

109

Desktop-native BigFish companion for DeepSeek Harness — real Agent status, always on top on Windows.

Installed cleanly when we ran it

npm packageJavaScriptyesterday

AI review

dsh-dafeiyu

What it is — This is a desktop-native companion plugin for DeepSeek Harness that displays real Agent status always on top in Windows.

Who it is for — If you are using DeepSeek Harness on Windows to handle projects that require observing the Agent's thinking process, this plugin provides real-time status display. Users running DeepSeek Harness on other operating systems do not need to install it because the plugin is optimized for Windows and WSL2 desktop display.

Watch out — It passes sandbox testing in a fresh profile and registers with harness. It executes shell commands but no obvious issues found.

The verdict — I would install it because it surfaces DeepSeek Harness real events in a always-on-top desktop window without requiring extra applications.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Source

dsh-stock-watch

Awu12277

44

A-share watchlist real-time market monitoring plugin: a collapsible popup in the top-right corner of the DeepSeek Harness (DSH) web interface for real-time quote monitoring, group switching, intraday and candlestick (K-line) charts, and buy/sell target price settings.

Installed cleanly when we ran it

npm packageJavaScriptyesterday

AI review

dsh-stock-watch

What it is — DeepSeek Harness web interface shows a collapsible popup in the top-right corner for real-time A-share watchlist monitoring.

Who it is for — If you need to monitor A-share watchlist real-time quotes, switch groups, and view intraday K-line charts in DeepSeek Harness using an agent for investment analysis research, this plugin suits you. If you do not want the plugin to execute shell commands, you can skip this plugin.

Watch out — Sandbox test passed: installed in a new profile and registered by Harness. No obvious pit found.

The verdict — I would install it because it runs charts in the browser and automatically injects DSH skills, but only if you accept the prompt that it will execute shell commands.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Source

deepseek-pet

keleus

29

在你的deepseek-harness上养一只吃白饭的大蓝鲸

Installed cleanly when we ran it

GitHub sourceJavaScriptyesterday

AI review

deepseek-pet

What it is — An interactive desktop pet plugin embedded in the DeepSeek Harness webpage, automatically switching expressions and playing animations.

Who it is for — Users performing analysis, answering, coding and network search in DeepSeek Harness are suitable to install this plugin because it switches expressions according to these scenarios and displays the session bar. In the environment of using only a single session without sub-agent in DeepSeek Harness, the parallel session display function of this plugin remains inactive.

Watch out — Sandbox test passed. The plugin was installed in a new profile and registered in the harness profile. No obvious issues found.

The verdict — I would install it because it prompts "eat more" when context occupation reaches 62%, providing interesting reminders during long-running tasks.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Source

dsh-emoji

hellodigua

24

Automatically add emojis to AI replies.

Installed cleanly when we ran it

npm packageTypeScriptyesterday

AI review

dsh-emoji

What it is — Add inline emoji support to DeepSeek Harness replies via semantic markers that render as images.

Who it is for — When using DeepSeek Harness models for customer service or creative replies that require tone or emotion, this plugin inserts markers like ::happy::. If you prefer pure text responses in conversations, this plugin is not needed.

Watch out — No obvious issues found. After installation in a fresh profile, it registers with harness. Custom ZIP uploads take effect on the next call.

The verdict — I would install it because it allows switching emoji styles within the same conversation to fit different expression needs.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
Source