AI Agent Hub
Back to plugins
🖥️

dsh-drop-path

Client Updated 2026.09.11

Run the following command in DeepSeek Harness:

dsh plugin install kongbai-shike/dsh-drop-path

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

Run dsh plugin install kongbai-shike/dsh-drop-path in your DeepSeek Harness terminal to install the plugin; the source code is available at https://github.com/kongbai-shike/dsh-drop-path .

About this plugin

DSH only accepts image attachments in its composer, so dropping a document, code file, or archive simply does nothing visible. dsh-drop-path intercepts those rejected drops, extracts the real absolute filesystem path from disk, writes it into the input box, and copies it to the clipboard, letting you hand the path straight to the model for reading.

Three working modes are available: Auto (default; non-image files get their path filled in while images still go through the normal attachment flow), Always (every file type produces a path), and Off (restore original drag behaviour). A single tap on the small toggle in the bottom-left corner switches modes, and the choice persists via localStorage. Under the hood the plugin uses the DSH Desktop preload bridge to resolve paths, listens at the window capture phase to stay clear of the built-in drag overlay, and includes a 1.2-second watchdog so the toast disappears on its own if the drag is cancelled.

Best for desktop users who frequently need a model to read local files such as PDFs, source code, config files, or logs. In a regular browser the plugin detects the missing local-path bridge and steps aside completely with zero side effects on DSH.

Use Cases

  • Drop a local PDF into DSH and hand the model a readable path
  • Grab a source or config file path instantly to feed the model as context
  • Copy an absolute filesystem path with a single drag-and-drop
  • Switch between three modes to control which drops get intercepted

Best For

  • DSH desktop users who frequently pass local file paths to the model
  • Developers working with PDFs, source code, and logs on disk
  • Efficiency-focused users who want a smoother drag-and-drop path workflow