dsh-markitdown
Run the following command in DeepSeek Harness:
dsh plugin install jiekesu967/dsh-markitdown
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install jiekesu967/dsh-markitdown in DeepSeek Harness to install; source at https://github.com/jiekesu967/dsh-markitdown
About this plugin
A text-only model hits a wall when it encounters a PDF, a spreadsheet, or a slide deck. Microsoft MarkItDown solves that conversion beautifully, but it is a Python package, and a harness plugin cannot assume the user has the right interpreter, the right install layout, or anything installed at all.
dsh-markitdown wraps MarkItDown as a single DeepSeek Harness tool. The model calls markitdown before reading any non-text file and gets back Markdown it can actually parse. The plugin probes four engines in order: the local markitdown CLI, a no-install uvx run, a Python module call, and a built-in zero-dependency converter. It works on a completely bare machine and still tells you exactly what to install for full fidelity. Size caps, subprocess isolation, archive decompression limits, and honest truncation announcements keep untrusted input from blowing up the pipeline.
Built for developers whose workflows regularly hand a model non-text documents: meeting records, data reports, customer PDFs, quarterly slide decks. The model simply calls the tool, reads the result, and moves on. No manual conversion, no Python environment to maintain.
Use Cases
- Read a client PDF contract and extract key clauses on the fly
- Convert quarterly Excel reports to Markdown for model-powered analysis
- Turn PowerPoint meeting notes into text so the model drafts action items
Best For
- Developers who regularly feed non-text files to an LLM and want zero manual conversion
- Technical users who need document ingestion without maintaining a Python stack
- Teams building DeepSeek Harness workflows where the model must auto-read files
Related Plugins
A unified suite combining hot runtime injection, task-aware thinking-mode routing, and a graded session protocol with red-team gates to sustain model diligence across long-horizon inference.
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.