When building DeepSeek Harness (DSH) agents, if you need to process structured medical data—for example, disease-drug associations or complication chains—simple textual memory is often insufficient. Directly integrating a complete medical knowledge graph is a more efficient solution. This article introduces the xby-medical-knowledge-graph plugin, which converts entities and relationships in the medical domain into graph database structures.

Plugin Positioning

The plugin is maintained by xby-skill and is a medical knowledge base plugin based on a graph database. It builds a data model around diseases, drugs, symptoms, diagnoses, complications, diet, and their relationships, for preliminary research scenarios.

Core Capabilities

The plugin provides the following capabilities:

  • set_xby_apikey: Sets the API key in the chat. This key is persisted automatically and remains valid after the DSH process is restarted.
  • get_schema: Retrieves the schema of the graph database, including node types, properties, and the connections between them.
  • execute_read_cypher: An MCP tool that allows read-only Cypher queries to be executed on the graph database.

Installation and Activation

The installation command requires the --profile web option. Run the following command to install it:

dsh plugin --profile web add github:xby_skill/xby-medical-knowledge-graph

Configuration and Usage

Before use, you need to obtain an API key. Please visit the Xiaobenyang Official Website to register and obtain the key, then call set_xby_apikey in the conversation to configure it.

The plugin is released under the MIT License. It is currently intended for preliminary research only, and specific use cases should be adjusted based on actual requirements.

Conclusion

This plugin provides DSH with structured medical knowledge retrieval capabilities.