AI Agent Hub
Back to plugins
🖥️

dsh-prompt-enhance

Client Updated 2026.08.17

Run the following command in DeepSeek Harness:

dsh plugin install BBbangage/dsh-prompt-enhance

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

Run dsh plugin install BBbangage/dsh-prompt-enhance in the DeepSeek Harness terminal to install; source code is at https://github.com/BBbangage/dsh-prompt-enhance .

About this plugin

Throwing a one-line request at a model in DeepSeek Harness usually yields a generic answer. dsh-prompt-enhance turns the sparkle icon next to the send button into a one-click enhancer: your current draft is expanded into a structured prompt that keeps your original sentence intact, followed by two to four naturally listed requirements you can fine-tune line by line, without the boilerplate of Role-Goal-Output labels.

The enhance logic dispatches by intent--parameter tuning, data aggregation, content creation, translation, review, code, file handling--each with its own emphasis, and auto-extracts data sources, formats, audiences, and timelines already present in your input so nothing is asked twice. By default it routes through your configured model for domain-specific rewrites; when the model is unavailable or times out at fifteen seconds, it silently falls back to a built-in rule template and still delivers a usable result. After enhancement the same button becomes an undo icon, restoring the original text in one click.

Built for developers, content operators, and data analysts who write prompts in DSH daily and want sharper output without spending ten minutes crafting them. Zero extra configuration, no separate API key, ready to use the moment you install it.

Use Cases

  • Click the sparkle icon to expand a vague one-liner into a structured prompt
  • Hit the undo icon to instantly restore your original draft after a bad enhance
  • LLM unavailable or timed out? Rule templates kick in and still deliver a result

Best For

  • Developers writing prompts in DSH daily who skip manual Role-Goal-Output crafting
  • Operators and content staff who want sharper output without ten minutes of prompt polishing
  • Analysts who need structured format, audience, and timeline requirements before running queries