Foreword¶

In agent development based on DeepSeek Harness (DSH), accurately referring to domain-specific terminology is key to ensuring answer quality. General knowledge bases often cannot cover all specialized terms in niche fields, or they may fail to update in a timely manner. A localized plugin solution can solve this problem.

Plugin Overview¶

The plugin is maintained by uckkk and is used for locally storing and retrieving terminology related to brain-computer interfaces. It does not depend on network requests and runs directly within the DSH process, making it suitable for developers who need a stable source of BCI (brain-computer interface) definitions.

Core Features¶

The plugin provides two tools, one for listing and one for detailed queries.

• list_bci: Lists knowledge entries for brain-computer interface terminology. The returned results include names and corresponding English terms.
• get_bci: Queries details of a brain-computer interface terminology entry. Supports retrieval by id or name.

Installation and Enablement¶

Install it using DSH plugin management commands. Ensure the current environment has the corresponding peer dependencies configured (@deepseek-ai/cordis and @deepseek-ai/dsh-tools).

dsh plugin add github:uckkk/dsh-brain-computer-terms

Usage¶

After successful installation, the plugin registers the tools described above in the current session. You can call them directly to obtain results.

Notes¶

• The plugin is a pure Node.js implementation and has no network dependencies at runtime.
• It is recommended to check the source code and license (MIT) before installation.
• It is suitable for use in scenarios involving discussions or development related to brain-computer interface technology.