dsh-llm-databricks
Run the following command in DeepSeek Harness:
dsh plugin install zhuchengde0214-ctrl/dsh-llm-databricks
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install zhuchengde0214-ctrl/dsh-llm-databricks inside your DeepSeek Harness workspace to install the bundle; the source lives at https://github.com/zhuchengde0214-ctrl/dsh-llm-databricks .
About this plugin
If you have conversation models—Claude, GPT, Gemini, Llama, DeepSeek, or others—already served behind a Databricks AI Gateway and want to call them from DeepSeek Harness the same way you would a local endpoint, dsh-llm-databricks was built for exactly that. It installs as a fully standalone bundle, touches neither the Harness core nor node_modules, and plugs in exclusively through the public LLM, settings, and credentials seams, so you are up and running the moment configuration is in place.
Use Cases
- Your Databricks Workspace already serves Claude, GPT, or DeepSeek via AI Gateway and you want to call them from Harness without patching core or node_modules
- A single workspace exposes MLflow Chat Completions, OpenAI Responses, and Anthropic Messages simultaneously; declare one provider route per protocol to unify access
- You need the bearer token kept strictly in the credentials service—never in settings files or logs—while still hitting HTTPS-gated endpoints
Best For
- Databricks Workspace administrators who have already wired up AI Gateway model endpoints
- Application teams running DeepSeek Harness that want cloud-served models without managing local inference
- Platform engineers with strict security, audit, and credential-isolation requirements for LLM call chains
Related Plugins
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.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.