In the DeepSeek Harness (DSH) ecosystem, developers often need to add specific calculation capabilities to agents. The binomial distribution is a common probability model used to calculate the probability of exactly k successes in n independent trials. The dsh-binomial-pmf plugin provides a local implementation of this mathematical function, requiring no network connection.

Plugin Overview

dsh-binomial-pmf is a DSH plugin maintained by uckkk. It provides binomial distribution probability mass function calculations and belongs to the model inference tool category.

Core Features

  • Pure Node implementation with no network dependencies
  • Provides a PMF (probability mass function) calculation interface

Installation and Enabling

Install using the official command:

dsh plugin add github:uckkk/dsh-binomial-pmf

Before installation, ensure the environment meets the dependency requirements:

{
  "peerDependencies": {
    "@deepseek-ai/cordis": "^4.0.1",
    "@deepseek-ai/dsh-tools": ">=0.1.0-rc.6"
  }
}

Typical Use

After integrating the plugin, PMF calculation logic can be invoked through DSH tools.

Use Cases and Cautions

This plugin is suitable for DSH bot or agent development scenarios that require local probability calculations and have network isolation requirements.

Cautions:
1. The plugin runs with the permissions of the current DSH process.
2. Check the source code before use to ensure it complies with security requirements.
3. The license is MIT.

Summary

The plugin solves local binomial distribution calculation needs through a pure Node implementation, making it suitable for scenarios that require deterministic calculation results. Directory address: https://www.skillhub.cn/plugins/uckkk/dsh-binomial-pmf, Source code address: https://github.com/uckkk/dsh-binomial-pmf.