self-evolving-router-dsh
Run the following command in DeepSeek Harness:
dsh plugin install sra-research/self-evolving-router-dsh
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install sra-research/self-evolving-router-dsh in your terminal to add the plugin; the source repository is available at https://github.com/sra-research/self-evolving-router-dsh .
About this plugin
In a DeepSeek Harness model-inference pipeline, routing decides which model instance or backend each request is sent to. self-evolving-router-dsh is an inference-routing plugin built for exactly this concern: instead of treating the routing policy as a fixed, hand-written configuration, it adopts a self-evolving approach that lets the routing logic adapt over time based on runtime feedback signals.
If you are building a multi-model or multi-backend inference service on DeepSeek Harness and want your routing decisions to evolve in response to load, latency, or quality signals rather than requiring constant manual re-tuning, this plugin offers a lightweight, MIT-licensed path forward. Maintained by the sra-research team, it stays focused on the inference-scheduling niche and suits engineering teams already working within the Harness ecosystem who need more dynamic routing behavior.
Use Cases
- Dynamically adjusting inference routing based on real-time load across multiple models
- Replacing static routing configurations to reduce manual tuning in inference pipelines
- Introducing adaptive scheduling strategies for inference requests in DeepSeek Harness
Best For
- Engineering teams building inference services on DeepSeek Harness
- Developers needing multi-backend dynamic routing without manual rule upkeep
- Platform architects prioritizing flexible and continuously evolving inference scheduling
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.