dsh-reasoning-effort-hdbzq

In DeepSeek Harness, the UI for model selection and reasoning effort adjustment can sometimes be less intuitive, hindering user efficiency. The dsh-reasoning-effort-hdbzq plugin solves this by replacing the original widget with a model list and a three-tier reasoning effort slider, offering a more flexible and responsive interaction. Its core capabilities include smooth slider operations like continuous tracking, elastic overshoot, and snap-to-position, providing immediate feedback when adjusting reasoning levels, while fully complying with the official plugin contract. This plugin is ideal for advanced DeepSeek Harness users, such as developers and researchers, who need fine-grained control over the reasoning process to optimize model performance or conduct experimental explorations.