dsh-promptkit
Run the following command in DeepSeek Harness:
dsh plugin install fsrmqi/dsh-promptkit
Paste the following prompt into your AI chat to install this plugin:
Fetch the plugin source from https://github.com/fsrmqi/dsh-promptkit and run dsh plugin install fsrmqi/dsh-promptkit to complete the installation inside DeepSeek Harness.
About this plugin
Prompt drafts are often vague and loosely structured, yet you still want them polished without the model drifting off-topic or inventing deliverables. dsh-promptkit sits inside DeepSeek Harness and gives you a one-click ✦ Enhance button: your draft is refined into a clearer, more actionable prompt in the preview panel while strictly preserving the original intent and scope. Semantic enhancement reuses the model already configured in DSH—no extra API key needed—and a local lightweight mode works entirely offline with zero telemetry.
Core capabilities include a library of 22 thinking methods (Socratic questioning, first principles, steel-man, and more) that are recommended rather than imposed; a local PromptKit Vault for searching, grouping, version-diffing, and exporting drafts; an optional five-dimension diagnosis (clarity, hidden premises, falsifiability, actionability, context fit); and three strength levels that control wording detail only, never task scope.
It is built for developers and AI engineers who write and iterate prompts frequently inside DSH, want a structured pre-send review, need a local library of reusable prompt assets, and care about data privacy with zero telemetry and local-only storage.
Use Cases
- One-click enhancement of raw prompt drafts into structured actionable prompts
- Matching one of 22 thinking methods to refine prompt reasoning
- Building and versioning a local prompt vault for reuse
Best For
- Developers who write and iterate prompts frequently in DSH
- AI engineers who need a structured pre-send review and value data privacy
- Team leads building a reusable prompt asset library
Related Plugins
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.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.