dsh-btw
Run the following command in DeepSeek Harness:
dsh plugin install loster12520/dsh-btw
Paste the following prompt into your AI chat to install this plugin:
To install the dsh-btw plugin, visit the source repository at https://github.com/loster12520/dsh-btw and then run the command dsh plugin install loster12520/dsh-btw in DeepSeek Harness.
About this plugin
In deep conversation systems, users often need to query information or run tests temporarily without disrupting the main dialogue context or consuming main model resources. The dsh-btw plugin is a bypass Q&A tool designed for this purpose: it creates an independent read-only session branch via the /btw command to handle questions. This branch has restricted tool permissions (such as file reading, searching, web queries), ensuring no system modifications, while Q&A content remains fully isolated from the main dialogue's LLM context, keeping the primary workflow clean.
Core capabilities include support for multi-turn dialogue, automatic synchronization of main dialogue history (injecting incremental information with each query), and per-session isolation to prevent cross-contamination between different conversations. The plugin presents the Q&A process in a centered pop-up panel, offering a convenient interaction experience. It is ideal for developers, testers, or anyone needing to quickly fetch additional information within a dialogue without interrupting the main task flow, helping users manage context efficiently and enhance productivity.
Use Cases
- Temporarily query information within the main dialogue without affecting context.
- Conduct multi-turn test Q&A while keeping the main dialogue clean.
- View independent answers via a centered pop-up panel to avoid disrupting the main flow.
Best For
- Developers for quick tests or information queries.
- Testers to verify model behavior without polluting the production environment.
- Dialogue system users needing temporary bypass functionality.
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.