Plugin directory / Desktop / deepseek-harness-desktop
deepseek-harness-desktop
Verified · install-tested on dsh MichengAI
What it does
Cross-platform desktop for DeepSeek Harness, no pre-installed environment required
Works — verified, early-stage project
Cross-platform desktop for DeepSeek Harness, no pre-installed environment required It installs cleanly and boots without issues in our testing. It's early-stage but functional.
“Verified” means our automated CI actually ran dsh plugin add in a clean profile and it booted — nothing more. Feature descriptions and version compatibility are the author’s claims. This is not a security audit and not an endorsement of third-party code.
README
DeepSeek Harness Desktop
A native desktop launcher for DeepSeek Harness on Windows, macOS, and Linux.
English · 简体中文
DeepSeek Harness Desktop is a community-maintained distribution project. It is not an official DeepSeek AI product.
DeepSeek Harness Desktop packages the DeepSeek Harness web experience as a native desktop application. It bundles the required Node.js runtime, starts DSH locally, and opens the interface in a dedicated desktop window.
Download
Download supported packages from GitHub Releases.
| Platform | Package | Release status |
|---|---|---|
| Windows x64 | .exe installer | Available · Download |
| Windows x64 | .zip archive | Available · Download |
| macOS Apple Silicon | .dmg installer | Available · Download |
| macOS Intel | .dmg installer | Available · Download |
| Linux x64 | .AppImage | Available · Download |
| Debian / Ubuntu x64 | .deb package | Available · Download |
All current and historical packages are available from GitHub Releases.
Screenshots
Start a new workspace conversation with model and permission controls.
Choose a built-in Agent preset or create a custom preset.
What it provides
- Native startup — installs and launches DeepSeek Harness on Windows, macOS, and Linux without a separate Node.js installation.
- Local DSH runtime — starts DSH on a randomly assigned loopback port and renders it in an Electron window.
- Focused desktop workflow — uses the existing DSH workspace, session, model, plugin, and Agent preset capabilities.
- Desktop behavior — single-instance protection, system-tray controls, and safe handling of external links.
- Continuity with DSH CLI — reuses the current user's DSH home directory, including existing settings and sessions.
Requirements
| Item | Requirement |
|---|---|
| Operating system | Windows 10/11, macOS, or Linux |
| Architecture | Windows x64, macOS Apple Silicon (arm64), macOS Intel (x64), or Linux x64 |
| Network | Required only for the model providers and tools you configure |
| Node.js | Not required for end users; bundled with the application |
Install and get started
- Download the package for your operating system from Releases.
- Run the Windows .exe installer, open the macOS .dmg, run the Linux .AppImage, or install the Debian/Ubuntu .deb package; then launch DeepSeek Harness Desktop.
- Wait for the local DSH service to start, then create or open a workspace and session.
- Configure models, plugins, permissions, and Agent presets in DSH settings as needed.
The installer bundles a complete runtime. Installation and first startup can take several minutes on some machines; allow the process to finish before cancelling.
Privacy and security
- The launcher accepts and loads only validated local HTTP addresses on 127.0.0.1.
- External HTTP(S) links open in the system browser; file, JavaScript, and data URLs are blocked.
- Electron runs with Node.js integration disabled, context isolation enabled, and sandboxing enabled.
- DSH configuration, sessions, and credentials remain in ~/.dsh (on Windows: %USERPROFILE%.dsh). Uninstalling the desktop launcher does not remove that directory.
- The launcher does not expose Electron IPC APIs to the DSH page.
Your configured model providers and DSH tools may make their own network requests. Review their respective settings and policies before use.
Troubleshooting
| Situation | What to do |
|---|---|
| Installation appears to pause | Wait several minutes. The installer is extracting the bundled runtime. |
| macOS says the app cannot be opened | Move the application to Applications, then open it again and follow the system prompt. |
| Startup fails | Reopen the app and check the diagnostic path shown in the error window. |
| Existing DSH data is not visible | Confirm you are using the same account and inspect ~/.dsh (on Windows: %USERPROFILE%.dsh). |
Development
Development requires Windows, Node.js 24.19.0, pnpm 11.20.0, and a built checkout of DeepSeek Harness.
[Console]::OutputEncoding = [System.Text.Encoding]::UTF8
$OutputEncoding = [System.Text.Encoding]::UTF8
pnpm install --frozen-lockfile
pnpm test
$env:DSH_RUNTIME_ROOT = 'D:\Repository\deepseek-harness'
pnpm run dist
Build output is written locally to *release* and is not committed. Pushing a vX.Y.Z tag builds Windows x64, macOS Apple Silicon/Intel, and Linux x64 AppImage/DEB packages.
Project documentation
Project status, active work, architecture constraints, and iteration records are available from the documentation entry point.
Support and contribution
- Report defects and feature requests through GitHub Issues.
- Before submitting a pull request, run the test suite and keep user-facing text in Simplified Chinese where applicable.
- See the upstream DeepSeek Harness repository for DSH-specific functionality and documentation.
License
This project is licensed under the Apache License 2.0.
Install
Install the catalog once, then DeepSeek Harness can find and install any plugin from this site automatically:
dsh plugin add dshbase-catalog Then say "install deepseek-harness-desktop for me" — your agent finds it in the directory and installs it. Docs: dshbase-catalog · verified packs.
This plugin is GitHub source (not published to npm) — install it straight from the repo:
Web profile:
dsh plugin --profile web add github:MichengAI/deepseek-harness-desktop Headless (CLI) profile:
dsh plugin --profile headless add github:MichengAI/deepseek-harness-desktop Test report
Verified: L1 install + L2 load + L3 runtime from GitHub source on dsh 0.1.0-rc.6.
When to use it
Bring a new model, provider, or routing policy into the loop so dsh can pick the right brain for the job.
Who it's for
Users juggling multiple models or providers who want cost, quality, and latency balanced automatically.
For developers — extending it
Provider adapters and routing heuristics are the seams — add a backend, tune the fallback chain, or add per-task model selection.