Introduction¶
DeepSeek Harness (DSH) adopts a plugin-based architecture. When an agent needs to handle specific business logic (such as OKR objective decomposition), using a lightweight plugin is more efficient than writing custom code or relying on external APIs. dsh-okr-split is a tool of exactly this kind, focused on local computation.
What It Is¶
The plugin is maintained by uckkk, and its core function is to provide the okr_split tool. It can decompose OKR objectives locally, perform precise calculations based on parameters, and return detailed results. It is implemented in pure Node.js and does not rely on network requests.
Core Features¶
- OKR objective decomposition: Supports decomposition calculations for objectives.
- Precise calculation: Performs calculations based on input parameters and returns detailed results.
- Local execution: Pure Node implementation with no network dependency.
- Tool integration: Registers the
okr_splittool for invocation in sessions.
Installation and Enablement¶
Install it using the officially provided command:
dsh plugin add github:uckkk/dsh-okr-split
After installation, the plugin registers the corresponding tool, which can then be invoked in the current DSH session.
Typical Usage¶
Invoke the okr_split tool directly in a session. The specific parameters should be filled in according to the required OKR decomposition logic, and the tool will return detailed calculated results.
Applicable Scenarios and Notes¶
- Applicable scenarios: DSH agent developers who need local OKR decomposition logic.
- Runtime environment: The plugin runs with the permissions of the current DSH process and depends on
@deepseek-ai/cordisand@deepseek-ai/dsh-tools. - Notes: It is recommended to review the source code before use to confirm the logic. Note that the license is MIT.
Conclusion¶
With dsh-okr-split, you can complete OKR objective decomposition tasks without increasing network overhead. For more details, see the project source code: https://github.com/uckkk/dsh-okr-split