prompt-optimizer-plugin
Run the following command in DeepSeek Harness:
dsh plugin install naitoupi/prompt-optimizer-plugin
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install naitoupi/prompt-optimizer-plugin in your terminal to add the plugin to DeepSeek Harness; the source is hosted at https://github.com/naitoupi/prompt-optimizer-plugin.
About this plugin
Polishing a prompt before hitting send can feel just as tedious as writing the actual code. You have a rough idea, type it out, revise it, and it still is not specific enough for the AI to deliver a useful answer. The Prompt Optimizer plugin closes that gap: a single Optimize button sits right next to the DSH input box, rewriting your rough draft into a clear, well-structured prompt in one click so you can move on to the actual task.
The workflow is deliberately minimal. Click the button and the input field instantly shows the polished version, with the original text floating above for a quick side-by-side check. Not happy? Hit undo and everything reverts cleanly with no trace left behind. Under the hood, the optimization is a standalone model call that never pollutes your ongoing conversation thread. In the settings panel you can tune thinking intensity, temperature, and max output length, or even edit the raw optimization instructions the model receives, with changes taking effect immediately and no restart required.
If you would rather spend that five-minute prompt-polishing cycle on the actual task and still want the AI to understand your intent the first time, this plugin fills the gap between jotting down a rough note and producing a ready-to-send prompt. The interface follows the DSH language automatically, serving both Chinese and English users equally, all under the permissive MIT license.
Use Cases
- Quickly rewriting a vague draft into a structured prompt inside a DSH conversation before sending
- Comparing the original and optimized versions side by side, with one-click undo if unsatisfied
- Calling the model in isolation to polish a prompt without polluting the existing chat context
Best For
- Users who want the AI to understand their intent on the first try and skip manual prompt polishing
- Developers and product professionals using DeepSeek Harness daily who value speed
- Individual developers who prefer bilingual UI, local-only settings storage, and the MIT license
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.