Introduction

When handling legal or financial issues in DeepSeek Harness (DSH) sessions, precise calculation results are often required. Traditional approaches usually rely on web searches or external tools, which not only increases network latency but may also involve the transmission of privacy-sensitive data. DSH’s plugin architecture allows capabilities to be extended locally. dsh-severance-pay is specifically designed to solve the problem of locally and accurately calculating severance pay.

What Is This

dsh-severance-pay is a client-side plugin maintained by uckkk. It focuses on the calculation logic for severance pay and provides a standard DSH tool interface, allowing agents or users to obtain calculation results directly in local sessions.

Core Features

  1. Parameterized Accurate Calculation: Supports calculating severance pay based on specific parameters (such as years of service and wage base) and returns detailed results with explanations.
  2. Pure Node Implementation: Developed with Node.js, without external network requests; runtime speed depends on local computational resources.

Installation and Activation

Run the following command in the terminal to install the plugin.

dsh plugin add github:uckkk/dsh-severance-pay

After installation, the plugin registers a tool named severance_pay in DSH sessions.

Typical Usage

After installation, simply call the tool registered by the plugin in your conversation. Users only need to provide input according to the parameter format defined by the tool, and the system will return the calculation results.

Applicable Scenarios and Notes

  • Applicable Scenarios: Suitable for scenarios requiring quick estimation or discussion of severance pay calculations under labor law in local sessions.
  • Notes:
    • The plugin uses a pure Node implementation and has no network dependencies.
    • Please ensure that the required dependencies are installed: @deepseek-ai/cordis (^4.0.1) and @deepseek-ai/dsh-tools (>=0.1.0-rc.6).
    • The code is open source and licensed under the MIT License. It is recommended to review the source code and license before use.

Links

Source code and documentation can be found at:
* GitHub Repository
* DSH Directory Page