dsh-provider-model-configurator

Configuring model entries for multiple providers in DeepSeek Harness can be tedious and error-prone. This plugin centralizes model management into the Model Pro settings page, where you can inspect a provider's explicit model entries and create, edit, duplicate-fill, or delete them without manually cross-checking configuration files.

You can configure context window, maximum output, input modalities, reasoning levels, and compatibility switches, while leaving fields unset to inherit catalog values instead of forcing explicit defaults. With the built-in model list and preset selector, you can quickly pull common context and max-token values or copy settings from a preset or another provider.

It is best suited to DeepSeek Harness administrators and deployers who maintain models across multiple providers and need a clearer way to manage model capabilities, reasoning behavior, and compatibility for inference dispatch.