AI Agent Hub
Back to plugins
🖥️

dsh-paste-path

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install jhuanxx44/dsh-paste-path

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

Run dsh plugin install jhuanxx44/dsh-paste-path in the DeepSeek Harness terminal to install the plugin, or grab it from the source repository at https://github.com/jhuanxx44/dsh-paste-path for a local install, then restart dsh web and refresh the page.

About this plugin

Copying a file in Finder with Cmd+C stores a native macOS file-list object on the clipboard, not a plain absolute path. Browser editors are intentionally isolated from the local filesystem, so handing a file location to an agent still means switching to a terminal and typing the path by hand—slow and error-prone.

dsh-paste-path closes that gap in one keystroke. It listens for Ctrl+V in the DSH composer, reads the file-list data from the macOS Pasteboard via osascript, and inserts the absolute paths into the current draft, one per line. Normal Cmd+V text paste is left completely untouched, the shortcut hint stays hidden when the clipboard holds no file paths, and duplicate paths already in the draft are skipped. No files are uploaded, no browser clipboard permission prompt is triggered, and ordinary paste behaviour is unchanged.

If you work on macOS and regularly feed local scripts, config files, logs, or project assets into DeepSeek Harness Web, this plugin turns what used to be a manual copy-paste ritual into a single, reliable Ctrl+V.

Use Cases

  • Copy a file in Finder, then press Ctrl+V in DSH to insert its absolute path into the composer
  • Hand a local script, config, or log path to an agent without retyping it in a terminal
  • Select multiple folders at once and insert each absolute path on its own line for the agent to read

Best For

  • macOS users who work in DeepSeek Harness Web day to day
  • Developers who reference local file paths frequently
  • Engineers who feed local project files into an agent workflow