AI Agent Hub
Back to plugins
⚙️

dsh-blanch-skill

Workflow Updated 2026.08.21

Run the following command in DeepSeek Harness:

dsh plugin install uckkk/dsh-blanch-skill

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install uckkk/dsh-blanch-skill in the DeepSeek Harness terminal to install the blanching and deodorizing plugin, available at https://github.com/uckkk/dsh-blanch-skill, then invoke the blanch_skill tool in your session to use it.

About this plugin

Blanching to remove gamy or fishy odors is one of the most error-prone steps in home cooking: cold-start versus hot-start, how long to blanch, whether to add ginger, scallion, or cooking wine, and exactly when to skim the foam all depend on the ingredient. dsh-blanch-skill turns that scattered kitchen experience into a single parameterized tool call, giving the cooking-knowledge stage of an agent workflow a reliable source of specific steps.

The plugin registers a blanch_skill tool in the session. Pass in the ingredient type, portion, and preference parameters, and it returns a structured blanching plan covering entry method, timing milestones, and deodorizing-aid pairings. The implementation is pure Node.js with zero external network dependencies, so it runs reliably in offline or air-gapped environments.

It is suited for developers building agent workflows that include a cooking-knowledge stage, and for users who want their AI assistant to give concrete step-by-step blanching instructions rather than vague general advice. The tool has a small, well-defined interface that makes it easy to chain with other ingredient-preparation or seasoning plugins in a full recipe pipeline.

Use Cases

  • Invoke in an agent workflow so the AI assistant returns concrete blanching steps per ingredient
  • Provide a structured deodorizing plan for a recipe pipeline in offline or air-gapped environments
  • Chain the blanching tool with ingredient-preparation and seasoning plugins into a full recipe flow

Best For

  • Developers building agent workflows with a cooking-knowledge stage
  • Users who want concrete executable blanching steps instead of vague general advice
  • Engineers deploying cooking toolchains in offline or air-gapped environments