AI Agent Hub
Back to plugins
🤖

dsh-plutonium

Model Inference Updated 2026.08.21

Run the following command in DeepSeek Harness:

dsh plugin install uckkk/dsh-plutonium

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install uckkk/dsh-plutonium inside DeepSeek Harness to install this plugin; the full source repository is at https://github.com/uckkk/dsh-plutonium.

About this plugin

Most model-inference stacks are tethered to a Python runtime, a CUDA driver, or a remote inference gateway, and a single misconfigured dependency can break the whole pipeline.

dsh-plutonium takes a deliberately lighter path: the inference logic is implemented entirely in Node.js and runs inside the Node process, eliminating cross-language calls and extra dependency chains. It also strips out every network request, so weights and computation stay local and the plugin works just as well with the Wi-Fi switched off.

This makes a natural fit for Node-ecosystem developers—whether you are shipping a CLI tool, a server-side API, or a browser-based offline app—who want to pull model inference in as an ordinary npm dependency instead of maintaining a separate Python runtime alongside it.

Use Cases

  • Running model inference locally in an air-gapped or offline environment
  • Embedding lightweight inference directly into a Node.js project
  • Building offline intelligent applications with zero external service dependencies

Best For

  • Node.js developers who want to skip a Python runtime
  • Tool authors deploying inference in air-gapped or offline networks
  • Frontend and full-stack engineers seeking a minimal dependency chain