dsh-iridium
Run the following command in DeepSeek Harness:
dsh plugin install uckkk/dsh-iridium
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install uckkk/dsh-iridium in your terminal to install the plugin from https://github.com/uckkk/dsh-iridium.
About this plugin
dsh-iridium tackles a deceptively common problem: running model inference without a network connection and without any native binary dependencies. It implements the entire inference path in pure Node.js, with no system-level C++ bindings and no outbound requests at runtime, so you can drop inference capability into a project the way you would any regular npm package.
The core offering is deliberately lean: a pure-Node implementation with zero network dependencies. That means it can run in air-gapped environments, CI pipelines, edge devices, or stripped-down Docker images, skipping the usual hassle of pulling in runtime libraries and configuring OS-level dependencies. Its MIT license lets you integrate it into commercial or open-source projects without friction.
Who is it for? If your project needs local inference but you want to avoid the platform-compatibility headaches of native addons, or you are assembling a fully offline inference stack, dsh-iridium gives you a light, clean starting point.
Use Cases
- Deploy model inference in a fully air-gapped environment
- Validate inference results locally inside a CI pipeline
- Run inference in minimal containers or on edge devices
Best For
- Node.js developers who need local inference without native addons
- Platform engineers building offline inference stacks
- Ops teams deploying AI capabilities in restricted-network environments
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.