Apache-2.0, TypeScript work system that runs natively on the embedded DeepSeek Harness runtime — local-first desktop coding plus self-hostable web agents and project spaces.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-08-30
License
Apache-2.0
01What can it help you accomplish?
Drive local coding work by conversation — write code, run commands, run tests, and change files in your own projects
A reviewable task view with agent activity, command output, test results, changed files, and diffs
Developers and teams who want coding work to stay on their local machine while keeping conversation, tool activity, and diffs together
Stand up shared project spaces, automation, and remote execution for a team
Shared task boards, files, discussions, execution status, and deliveries across people and devices
Teams that need to move a project forward across people, devices, or a self-hosted environment instead of leaving code on one computer
02How to install into DeepSeek Harness
Prerequisites
- Wegent Desktop is an Electron desktop app, so it needs a macOS / Windows / Linux desktop environment
- Self-hosting Wegent Web and Backend needs Docker plus a reachable model endpoint
Installation steps
- 01
Wegent Desktop: download the latest installer, install it, open the app, add or select a local project, then create a task and describe the work
- 02
Wegent Web and Backend (self-hosted): start the standalone Docker application with the one-line install script
Verify the integration
- For the self-hosted web app, open http://localhost:3000 and complete the setup flow (administrator password + model config)
03DSH integration and capability boundaries
Wegent is built directly on the embedded DeepSeek Harness (DSH) runtime — the Wework desktop workbench is composed from a set of DSH plugins, so DSH is the first-class application/plugin runtime rather than an external integration.
Desktop workbench
local project files, commands, and tests→a task view with agent activity, command output, test results, changed files, and diffs
Reusable agents
models, prompts, Skills, knowledge, tools, and collaboration settings→composable, reusable agent configurations
agent configs reference model, knowledge, and tool credentials configured for the workspaceMultiple execution targets
a coding task for a project→task execution on a local machine, a remote work machine, or a server-managed executor
tasks can run on remote machines or server-managed executors, driving code and commands across devicesSelf-hosting
Wegent Web and Backend deployment→team-facing web app with APIs, permissions, scheduling, and knowledge services
self-hosted deployment stores project data and model configuration on infrastructure you control
04Who is it for? When not to use it?
Good for
- Developers and teams who want coding work to stay on their local machine while keeping conversation, tool activity, and diffs together
- Teams that need to move a project forward across people, devices, or a self-hosted environment instead of leaving code on one computer
05Compatibility, maintenance and safety notes
- Building or developing Wegent from source requires Node.js 20+ and pnpm; ordinary use of the Desktop installer does not. The dependency is only a constraint for contributors and desktop developers.
- Wegent Desktop is an Electron desktop application and runs on a desktop OS (macOS / Windows / Linux). The self-hosted Web and Backend require Docker and a reachable model endpoint.
- The self-hosted web app needs network access to a model endpoint (model calls go to a remote model), and exposes http://localhost:3000 only for the initial setup flow.
Apache-2.0 · actively maintained (latest release wework-v0.3.3, 2026-08-30)
06Frequently asked questions
How does Wegent integrate with DeepSeek Harness?
DeepSeek Harness (DSH) is the embedded application and plugin runtime inside Wegent. The Wework desktop workbench is composed from a set of DSH plugins (dsh-app-wework, dsh-ui-*, dsh-electron-host, dsh-executor-runtime, dsh-terminal-runtime), so DSH is the first-class runtime rather than an external add-on. DSH itself does not run coding tasks; it discovers plugins and drives the Executor, which then drives Codex.
Do I need to install dsh separately?
No. The README describes DSH as the embedded runtime of the Wegent desktop app — it ships inside Wegent Desktop rather than as a separately installed CLI. You install Wegent Desktop (or self-host Web/Backend via Docker) and DSH is already present.
What is the difference between Wegent Desktop and Wegent Web?
Wegent Desktop is the local Electron workbench that works directly with your local projects, files, commands, and tests. Wegent Web provides the browser interface for remote agents, and Wegent Backend connects the desktop and web apps to shared project spaces, models, and execution devices.
How do I self-host Wegent?
Run the one-line Docker install in standalone mode: `curl -fsSL https://raw.githubusercontent.com/wecode-ai/Wegent/main/install.sh | bash -s -- --standalone`, then open http://localhost:3000 to set an administrator password and configure a model. Other deployment options are in the Installation Guide.
What are the platform requirements?
Wegent Desktop is an Electron desktop app for macOS / Windows / Linux. Building from source needs Node.js 20+ and pnpm. Self-hosting Web and Backend needs Docker plus a reachable model endpoint; the setup flow runs on http://localhost:3000.
07Related DSH workflows
awesome-gpt-image-2
by freestylefly
Prompt-as-Code engine and template library for GPT-Image2, with 470+ reverse-engineered cases, 20+ production templates and distilled agent skills.
voyager
by nagi-ovo
Browser extension that enhances Gemini, AI Studio, Claude and ChatGPT with a prompt manager that works across any web UI, DeepSeek Harness included.
dsh-im
by xmanrui
把 IM 机器人接入 DSH：统一管理飞书/微信/钉钉/企业微信/QQ/Telegram/Discord/WhatsApp 八个渠道，扫码或凭据接入、流式回复与设置页管理
chatgpt-gemini-timeline
by houyanchao
Chrome extension that adds folders, a prompt library, and a conversation timeline to ChatGPT & Gemini chat pages — works alongside DeepSeek Harness.
08Data and sources
DeepSeek Harness (DSH) is the embedded application and plugin runtime, and the Wework product UI is composed from a set…
DSH does not execute coding tasks itself. `dsh-app-wework` defines the product shell and extension points,
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-08-30. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.