AI Agent Hub
Back to plugins
dsh-virtual-workspace preview

dsh-virtual-workspace

Workflow Updated 2026.08.16

Run the following command in DeepSeek Harness:

dsh plugin install KevinWen7415/dsh-virtual-workspace

Paste the following prompt into your AI chat to install this plugin:

In DeepSeek Harness, use the command dsh plugin install KevinWen7415/dsh-virtual-workspace or provide the full GitHub URL https://github.com/KevinWen7415/dsh-virtual-workspace to the agent to install the plugin.

About this plugin

In DeepSeek Harness (DSH), agents are often confined to working within a single project directory, which becomes cumbersome when simultaneously handling multiple related codebases. The dsh-virtual-workspace plugin addresses this pain point by introducing virtual workspaces. It allows users to group multiple local directories into named virtual workspaces, enabling the agent to seamlessly read, search, and modify files across them as if they were under a single path. Core capabilities include providing 13 vws tool actions (such as list, add, read, write files), seamless integration with the DSH sidebar (via a drawer panel for management), automatic mirroring of workspaces to the built-in list, and sandbox-compliant write permission escalation. This plugin operates globally, supports both Chinese and English interfaces, and persists workspace definitions. It is especially suited for developers, data engineers, or anyone relying on DSH for complex code management who frequently needs to collaborate across multiple projects, significantly boosting agent efficiency and contextual coherence.

Screenshots

Use Cases

  • Search and fix code issues across multiple project directories
  • Create and manage virtual workspaces via the sidebar drawer
  • Read or modify files under virtual paths using vws tools

Best For

  • Software developers who need to handle multiple codebases simultaneously
  • AI researchers using DeepSeek Harness for complex project management
  • DSH users aiming to enhance agent cross-directory collaboration efficiency