dsh-model-manager: A Central Control Console for Local LLM Inference Services

`dsh-model-manager` is a plugin for DeepSeek Harness, designed to address the confusion in configuring multiple local LLM inference services. It provides a unified panel with support for service registration, parameter profile management, GPU detection, VRAM validation, and one-click benchmarking. The plugin safeguards security through a strict two-step confirmation mechanism and must be installed in a DSH environment to help developers efficiently manage local multi-model services.

Read More