jingyun-dsh: A Commercial Client That Turns DeepSeek Harness Agents into Sellable Products
jingyunstudio/jingyun-dsh
A Jingyun client plugin for DeepSeek Harness that adds membership login, subscription billing, compute quota control and an app marketplace, turning agents and skills into sellable products.
jingyun-dsh is a commercial desktop client built for DeepSeek Harness, aimed at turning agents, skills and workflows into products you can price and charge for. It fills the gap most builders hit: phone-number login, membership tiers, subscription and pay-as-you-go billing, compute quota control, plus cloud sync for chat history and artifact files. Its plugin library bundles 800+ community open-source plugins, with agents and skills listed and traded in one marketplace; brand name, logo and domain are replaceable through a config file, and everything can be packaged into a Windows desktop client with an installer.
How to Install
dsh plugin --profile web add @jingyun-ai/jingyun-dsh- Category
- Tools & Capabilities
- Platform
- DSH Plugin
- Author
- jingyunstudio
- Distribution
- Plugin
jingyun-dsh Key Features
jingyun-dsh Repo Summary
What Does It Do?
Jingyun DSH Client is a DSH plugin for DeepSeek Harness that turns agents, skills and workflows into sellable products. It solves the problem of AI builders having a working product but no way to charge for it, adding a full commercial loop inside the harness: sign-up and login, membership tiers, subscription payments, compute quota control and cloud asset sync. Maintained by jingyunstudio and written in TypeScript, it has around 541 stars and was last updated in September 2026.
Core Features
- Membership and subscriptions: phone-number login with subscription, pay-as-you-go and distribution models, plus a built-in compute quota system that tracks token usage and cost.
- Agent and skill libraries: an agent identity system that switches per session, alongside reusable skill packs that extend what the AI can do.
- Plugin library and app marketplace: over 800 community open-source plugins, with agents, skills and plugins listed and priced in one marketplace.
- Cloud data and artifact workbench: chat history, knowledge bases and output files sync to the cloud, while a resizable side panel previews code, documents and charts.
- Branding and desktop packaging: customize brand name, logo and domain through a config file, then package everything into a Windows desktop client with an installer.
How to Use This Plugin?
If you already run a DSH instance, connect the branding plugin, then copy and rename the example config file locally and fill in your own platform details to activate membership and payment features. Teams that need a standalone branded desktop client can pull the whole repository, install dependencies, build the plugin, and use the GUI packaging console to set the brand name, domain and tenant ID before producing a Windows installer. During development you can either start the DSH web service in a browser or preview the desktop shell in Tauri dev mode; packaging the desktop client requires preparing the embedded Node.js, Python and Git portable runtimes described in the README and placing them in the designated resource directory.
How to Troubleshoot This Plugin?
If packaging the desktop client reports missing runtime environment packages, that is expected: the embedded Node.js, Python and Git portable builds are large and not shipped in the repository, so obtain them as the official instructions describe and place them in the designated vendor resource directory. If your branding does not take effect, you most likely have not yet copied the example file into a real config file, or the domain, name and logo fields are incomplete; fill them in per the README branding section and rebuild.
This page is an independent rewrite of the plugin's official README — for authoritative documentation and the latest changes, refer to the source: jingyunstudio/jingyun-dsh. The plugin is third-party code that runs on your machine once installed; inclusion does not imply endorsement — please review the source before installing.