AI Agent Hub
Back to plugins
🤖

dsh-plugin-custom-provider-enhancer

Model Inference Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install cinob/dsh-plugin-custom-provider-enhancer

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

To install this plugin in DeepSeek Harness, run the installation command: dsh plugin install cinob/dsh-plugin-custom-provider-enhancer, with the source code located at: https://github.com/cinob/dsh-plugin-custom-provider-enhancer.

About this plugin

Connecting custom gateways like OneAPI or vLLM to DeepSeek Harness often results in raw model IDs without critical runtime parameters such as Context Window or Multimodal support, causing inaccurate token pressure measurement and context overflows. DSH Custom Provider Enhancer solves this by intercepting model discovery to automatically populate specifications like Context Window, Max Output Tokens, and Deep Thinking levels, resolving configuration gaps entirely.

Use Cases

  • Connecting third-party model gateways like OneAPI or NewAPI.
  • Enabling visual features such as image upload in the chat interface.
  • Resolving context overflow issues caused by missing model specifications.

Best For

  • Developers using DeepSeek Harness.
  • Users requiring precise control over context windows.
  • Operations staff managing multi-model API gateways.