dsh-tui-model-switch
Run the following command in DeepSeek Harness:
dsh plugin install FUSU123fusu/dsh-tui-model-switch
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install FUSU123fusu/dsh-tui-model-switch in the DeepSeek Harness terminal to install this plugin; source repo: https://github.com/FUSU123fusu/dsh-tui-model-switch .
About this plugin
In dsh-TUI, the main session and its subagents share a single model route by default. Yet in real workflows, subagent tasks - summarization, retrieval, formatting - are often lighter and do not justify the same costly provider or highest reasoning effort as the main conversation. dsh-tui-model-switch exists to change that: it lets you assign a dedicated provider, model, and reasoning level to subagents while leaving the main session route untouched.
Every setting is persisted in the dsh settings service under the model-switch namespace and edited directly through the TUI /settings screen in the Model Routing section, with live reload on change. You pick the subagent model from the configured provider catalog or type the provider/model manually, and optionally set reasoning effort (low / medium / high / max). At runtime the plugin wraps subagent start paths to merge the custom selection into agentOptions, and injects reasoning effort via the child agent request interceptor so it only affects subagent-origin sessions. Explicit provider or model fields set by the caller always take precedence over this setting.
If you already rely on dsh-TUI and dispatch subagents regularly, this plugin lets you tier costs and capabilities per task type without altering main-session behavior. Switching the subagent model to Follow Main Session restores the exact pre-install behavior. Non-TUI hosts continue to work as before: the command simply reports the active route as plain text and the settings section does not render.
Use Cases
- Run routine subagent tasks on a cheaper model while the main session stays on a premium provider
- Keep the main session at highest reasoning effort and lower it for subagent work to save cost
- Mix providers across main session and subagents, each on its own model route
Best For
- dsh-TUI developers who dispatch subagents regularly
- dsh users who want tiered model cost control per task type
- Configuration managers who prefer editing model routes in the TUI settings screen
Related Plugins
A service-oriented sidebar framework for DSH, offering a full workbench with file explorer, embedded browser, real terminal, Git panel, and extensible plugin services.
A beautiful, practical Claude Code-style TUI plugin with pixel whale top bar, flowing glow title, real-time status line, streaming thought expansion, time rewind, context progress bar and TPS gauge — zero core changes.
The plugin market for DeepSeek Harness: browse/search/one-click install community plugins and themes, with updates, backup/restore, hot disable, diagnostics, and AI fix.
A persistent whale widget on the DSH web UI that shows DeepSeek balance, today's usage, and per-turn cost, with drag-and-snap and sound effects.