In model inference or agent development workflows, generating periodic summaries (such as weekly reports) often involves tedious text organization. The core philosophy of DeepSeek Harness (DSH) is “everything is a plugin”; by extending the plugin system, features can be added without modifying core code. dsh-weekly-report is one of the tools in this ecosystem, designed to address the need for generating such periodic reports.

This is a plugin maintained by uckkk and categorized as model inference. It provides a structured way to generate weekly reports, with a core tool named weekly_report.

Core Features

The plugin provides the following capabilities:

  1. Structured method: Provides the weekly_report tool, which supports returning the structured method based on parameters.
  2. No network dependency: Pure Node implementation; no external network connection is required at runtime.

Installation and Activation

Add the plugin using the official installation command:

dsh plugin add github:uckkk/dsh-weekly-report

After installation, the plugin will register the corresponding tools in the current session.

Usage

You can directly invoke the tools registered by the plugin within a session. For specific parameters, refer to the plugin’s documentation definitions.

Technical Details

  • Dependency requirements:
    • @deepseek-ai/cordis: ^4.0.1
    • @deepseek-ai/dsh-tools: >=0.1.0-rc.6
  • License: MIT
  • Implementation: Pure Node.js implementation.

Conclusion

dsh-weekly-report is a lightweight tool focused on providing structured weekly report generation capabilities. The community directory has no official affiliation with DeepSeek or Huanfang, and the plugin is provided solely as an open-source tool. For more details, see the directory page or the GitHub repository.