Introduction¶
In the DeepSeek Harness (DSH) ecosystem, plugins are key to extending core capabilities. Developers often need clear configuration suggestions for specific scenarios. The dsh-basic-training plugin provides basic training functionality, can return clear suggestions based on parameters, and can run without an internet connection. It is suitable for scenarios with network-isolation requirements.
Plugin Overview¶
This plugin is maintained by developer uckkk and belongs to the model inference category. It is a DSH plugin implemented purely in Node.js, focusing on providing basic training-level suggestions.
Core Features¶
- Basic training suggestions: Returns clear training suggestions based on input parameters.
- Local execution: Implemented purely in Node, with no network dependency.
Installation and Activation¶
To install this plugin, run the install command in the DSH environment. After installation, the tools registered by the plugin will be available in the current session.
dsh plugin add github:uckkk/dsh-basic-training
Typical Usage¶
After installation, directly invoke the tool registered by the plugin in the session. The specific usage depends on the tool’s interface definition; the plugin itself provides entry points for calling basic training-related functions.
Notes¶
- Dependencies: The plugin depends on specific peerDependencies, including
@deepseek-ai/cordis(^4.0.1) and@deepseek-ai/dsh-tools(>=0.1.0-rc.6). Ensure that the environment versions meet the requirements. - Permissions and Security: The plugin runs with the permissions of the current DSH process. It is recommended to review the source code and license (MIT) before installation.
Summary¶
This plugin provides a lightweight, local training-suggestion solution that is suitable for integration into existing DSH workflows. For more details, refer to the plugin directory or the GitHub repository.