dsh-model-groups
Run the following command in DeepSeek Harness:
dsh plugin install roclee2692/dsh-model-groups
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install roclee2692/dsh-model-groups in DeepSeek Harness; the source is available at https://github.com/roclee2692/dsh-model-groups. Follow the README to add the entry to your profile bundles or profile patch after installation.
About this plugin
As the number of providers in a DSH model directory grows, the flat built-in list gets harder to navigate. dsh-model-groups solves this with a deliberately light approach: it overrides only the conversation.input.model slot and layers an expandable manufacturer grouping on top, so you can narrow down to DeepSeek, OpenAI, Anthropic, or any other vendor at a glance instead of scrolling through dozens of entries.
The grouping is purely presentational. Original provider IDs are preserved, routing for unknown models still falls back to Other/platform, and advanced controls such as Thinking are sourced from the host metadata of the selected model rather than a hardcoded capability set. The plugin handles no credentials, does not replace the community skin or marketplace, and renders its dialog in the browser native top layer to avoid clipping by theme containers. Uninstalling restores the built-in model picker immediately, and the /model command remains owned by the built-in plugin.
It suits users on DSH 0.1.1-rc.2 with a large model catalogue who want vendor-level filtering without pulling in a new skin or altering the marketplace. As an independent client extension that reuses DSH open-source slot and model-directory APIs, it does not guarantee compatibility with every future release; verifying in an isolated profile before upgrading is recommended.
Use Cases
- A large model catalogue makes the flat built-in list hard to scan
- Want vendor-level filtering without swapping out the community skin
- Reduce scrolling and memorization across many providers
Best For
- DSH 0.1.1-rc.2 users with large multi-provider model directories
- Client users who value native interactions and minimal changes
- Cautious users who want easy uninstall and an intact /model command
Related Plugins
A unified suite combining hot runtime injection, task-aware thinking-mode routing, and a graded session protocol with red-team gates to sustain model diligence across long-horizon inference.
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.