tirol-dsh-language-set
Run the following command in DeepSeek Harness:
dsh plugin install Emck/tirol-dsh-language-set
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install Emck/tirol-dsh-language-set in the DeepSeek Harness terminal to install the plugin; the source code is available at https://github.com/Emck/tirol-dsh-language-set , then restart the DSH service to activate it.
About this plugin
In a DeepSeek Harness workflow, the AI's reply language and internal thinking language are typically coupled, leaving users without a way to specify them independently. tirol-dsh-language-set decouples these two dimensions through system prompt injection so you can set the reply language and the thinking language separately; both currently support Chinese and English.
The plugin exposes seven configuration options covering the global default language, per-channel language overrides, custom prompt templates for replies and thinking, an optional language command registration, and a verbose logging toggle. All dependencies are dev-time only, so the build output carries zero runtime cost, and the plugin takes effect immediately after a service restart.
It is well suited for teams that need precise bilingual switching, for users who prefer thinking in one language while replying in another, and for DSH operators who want a quick in-service command to adjust language settings on the fly.
Use Cases
- Thinking in Chinese while replying to English-speaking users, or the reverse
- Switching the global reply language instantly via a single in-service command
- Customising prompt templates to fine-tune language style for replies and thinking separately
Best For
- Day-to-day DeepSeek Harness operators
- Developers who frequently alternate between Chinese and English workflows
- AI product teams that want thinking and replies in different languages
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.