AI Agent Hub
Back to plugins
🖥️

dsh-doc-quick

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install Entity-Him/dsh-doc-quick

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

Run dsh plugin install Entity-Him/dsh-doc-quick in DeepSeek Harness to install this plugin; the source code is available at https://github.com/Entity-Him/dsh-doc-quick

About this plugin

When you need an AI agent to work on a local document, sandbox boundaries often get in the way and add a layer of friction you did not sign up for. dsh-doc-quick turns that friction into a single drag-and-drop gesture: drop a file into the dsh web chat, the agent reads it straight from your local filesystem with no sandbox limits, and when the work is done a right-hand sidebar surfaces the outputs along with their file paths, ready to preview or download.

The plugin covers the document formats you actually work with day to day: PDF, Word (doc/docx), PowerPoint (ppt/pptx), EPUB, Jupyter Notebooks (ipynb), Markdown, and plain text. Under the hood the agent is equipped with three focused tools—doc_list to check pending tasks, doc_read to pull in file content, and doc_complete to register the produced artifacts—so you simply ask it to read or summarize without specifying paths or adjusting any configuration.

If your workflow involves regularly feeding local documents to an AI for reading, summarizing, or rewriting, dsh-doc-quick slots in naturally as a piece of the dsh web experience. The implementation is pure JavaScript with no native dependencies or Node-gyp build steps; PDF parsing lazily loads pdfjs-dist when available and degrades gracefully when it is not, so the setup is straightforward and ready to use the moment you open the chat.

Use Cases

  • Drag a PDF report into the chat and let the agent read it and produce a summary
  • Drop a Jupyter Notebook into the input and ask the agent to extract code and key conclusions
  • Feed a Word contract to the agent for clause extraction and path-based output

Best For

  • dsh web users who regularly need an AI to read local files
  • Developers who want to skip sandbox setup and process local files directly
  • Researchers and analysts working daily with PDF, Office, and Notebook formats