AI Agent Hub
Back to plugins
🖥️

dsh-reasoning-slider

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install qjcnmd/dsh-reasoning-slider

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

To install this plugin in DeepSeek Harness, run the command dsh plugin install qjcnmd/dsh-reasoning-slider, or visit the full source address https://github.com/qjcnmd/dsh-reasoning-slider for details.

About this plugin

When using DeepSeek Harness, users often need to switch between different reasoning tiers to balance response speed and generation quality, but the native interface may lack intuitive graphical controls, making the process cumbersome. The dsh-reasoning-slider plugin effectively addresses this by embedding a slider in the model selector, allowing users to easily drag and switch reasoning tiers from off to max, with real-time preview and automatic adaptation to model settings.

Its core capabilities include: automatically displaying the supported reasoning tiers for the selected model, instant switching on drag with immediate effect, and automatic tier carrying when switching models—if a model doesn't support the current tier, it reverts to its default. Additionally, it supports keyboard shortcuts (such as arrow keys) and scroll wheel interaction for enhanced convenience.

The dsh-reasoning-slider is particularly suitable for regular users of DeepSeek Harness, such as developers, researchers, and AI enthusiasts, who frequently adjust reasoning settings to optimize model performance in experiments or production environments. By simplifying the workflow, the plugin helps users focus more on content creation and task execution, thereby improving overall efficiency and user experience.

Use Cases

  • Quickly switch reasoning tiers in DeepSeek Harness to balance response speed and quality.
  • Experimentally compare the effects of different reasoning levels on model outputs.
  • Dynamically adjust model performance in production to adapt to varying task needs.

Best For

  • Regular users of DeepSeek Harness, such as developers and researchers.
  • AI practitioners requiring fine-grained control over model reasoning parameters.
  • Tech enthusiasts seeking efficient model interaction experiences.