dsh-ontology-kernel
Run the following command in DeepSeek Harness:
dsh plugin install aiko-dsh-plugins/dsh-ontology-kernel
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install aiko-dsh-plugins/dsh-ontology-kernel in the DeepSeek Harness terminal to fetch and install the ontology kernel bundle from https://github.com/aiko-dsh-plugins/dsh-ontology-kernel.
About this plugin
Building business plugins on DeepSeek Harness often means juggling ontology concerns—registering object types, link types, Functions, and Actions—while local persistence and browser-side projection live in separate, ad-hoc pieces. dsh-ontology-kernel collapses both into a single installable bundle that exposes one unified ontology interface, keeping adapter mechanics out of the way for upper-layer developers.
Two internal paths do the heavy lifting: business plugins contribute versioned object types, link types, Functions, and Actions through ctx.ontology.register(...), and the local adapter persists them via the Harness storage domain. When a front-end view is required, the remote adapter projects the active runtime to the browser through typed Typert Remotes, so browser plugins simply consume ctx.ontologyRemote without any dependency on the storage layer. The two adapters are inserted in a fixed order, and a later profile patch can replace either one without touching a single line of business code.
If you are writing a business plugin on Harness and need a structured ontology model to describe domain objects, their relationships, and executable actions—or if you want the front-end and back-end to share the same type definitions instead of maintaining parallel copies—this bundle is built for that. It does not opine on your domain logic; it simply guarantees a consistent, replaceable, and versioned registration path for every ontology definition.
Use Cases
- Register versioned object types, link types, Functions, and Actions inside a Harness business plugin
- Consume runtime ontology data from a browser plugin without depending on the local storage adapter
- Swap the local or remote adapter via a profile patch while keeping business code untouched
Best For
- Developers authoring business plugins on DeepSeek Harness
- Full-stack teams that need a single shared ontology type model across front-end and back-end
- Architects who want versioned ontology definitions with replaceable adapter implementations
Related Plugins
A unified suite combining hot runtime injection, task-aware thinking-mode routing, and a graded session protocol with red-team gates to sustain model diligence across long-horizon inference.
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.