AI Agent Hub
Back to plugins
dsh-attachments preview

dsh-attachments

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install WJZ-P/dsh-attachments

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

In DeepSeek Harness, you can install this plugin by running the command `dsh plugin install WJZ-P/dsh-attachments`, with the full source code available at: https://github.com/WJZ-P/dsh-attachments.

About this plugin

When using DeepSeek Harness for multimodal or code-centric conversations, we often need to provide local files or entire project folders as context to the model. While the native Harness excels at handling single images, managing numerous regular files—especially those requiring preserved directory structures—can be a fragmented experience involving manual uploads and lost hierarchy.

The dsh-attachments plugin is built to solve this exact pain point. It injects a robust and unobtrusive file system integration into Harness. Simply drag-and-drop an entire folder into the chat window, and the plugin intelligently converts it into a single attachment card while fully preserving its internal directory tree. Upon submission, files are persisted to a clear path within the workspace, enabling the model to access them as easily as if reading a local project. Best of all, it’s designed to be low-intrusive, seamlessly coexisting with Harness’s native image previews and message history, and remains invisible when no attachments are present.

This plugin is ideal for developers, researchers, and content creators who frequently engage in deep collaboration with AI models. If your workflow involves cross-file analysis, understanding project codebases, or providing complex repositories of documents and datasets all at once, dsh-attachments significantly boosts your efficiency and experience, making file management intuitive and orderly.

Screenshots

Use Cases

  • Drag an entire project folder to let the model quickly understand code structure and dependencies
  • Upload a set of research documents at once for cross-file analysis and Q&A
  • Submit a dataset with multi-level directories to the model, preserving its original organization

Best For

  • Developers who collaborate with AI to analyze complex codebases
  • Researchers frequently working with cross-referenced documents and data sources
  • Content creators looking to contextualize complete resource libraries for higher-quality generation