Introduction¶
The core philosophy of DeepSeek Harness (DSH) is “everything is a plugin.” When building agents, it may sometimes be necessary to integrate domain-specific computation logic instead of relying solely on the general-purpose reasoning of large language models. For the scenario of designing fat-loss meal plans, dsh-meal-plan provides parameter-based precise calculation capabilities.
What This Is¶
This is a DSH plugin maintained by developer uckkk, whose core function is “fat-loss meal planning.” It performs calculations in a parameterized way and provides detailed meal-planning results, making it suitable for scenarios that require directly invoking nutrition calculation logic within a conversational flow.
Core Features¶
The plugin registers a tool named meal_plan. This tool accepts specific parameters, performs fat-loss meal planning calculations, and returns detailed result data.
Installation and Enablement¶
Use the official installation command to add the plugin to the current environment:
dsh plugin add github:uckkk/dsh-meal-plan
After installation, the plugin becomes active, and the tool it registers can be invoked in the current session.
Typical Usage¶
According to the official documentation, after installation, you only need to directly invoke the tool registered by the plugin in the session. The tool supports precise calculations based on parameters and returns detailed results.
Applicable Scenarios and Notes¶
This plugin is suitable for agent development scenarios that require nutritional meal planning calculations for fat loss. Its implementation is based on pure Node.js, does not depend on network requests, and has relatively high execution efficiency. Note that the plugin runs with the permissions of the current DSH process. Before installation, it is recommended to review the source code to ensure it meets security and licensing requirements.
Conclusion¶
This tool adds a lightweight calculation module to the DSH plugin ecosystem. For more details, refer to the plugin directory and source code repository.