Global Simplified-Chinese mode: a 中 switch in the input box that injects a Chinese-language requirement into every session's system prompt when enabled, skipping anchored presets' pre-promotion round.
Install
# from GitHub (first run asks for allowBuilds approval — follow the hint, retry)
dsh plugin --profile web add github:dawnliming/dsh-chinese-mode
Any plugin you install runs third-party code with your own permissions — it can read your files, use your credentials, and reach the network, and tool approvals don’t sandbox it. GitHub-sourced plugins also run build scripts at install time. Only install sources you trust, and pin a commit (github:owner/repo#sha).
README
Global Chinese-mode toggle for DeepSeek Harness Web. A small 中 switch next to the permission selector in the input box; when enabled, a Chinese-language requirement is injected into the system prompt of every session preset.
English is provided for international users. The default documentation is Simplified Chinese (see README.md).
Features
- Global, persisted — the switch lives in host-side settings (
dsh-chinese-modenamespace) and applies to all sessions. - Unconditional injection — when the switch is on, the Chinese requirement is always injected (inserted just before the persona section), no "already has Chinese" detection.
- Works with every preset, including complete personas.
- Anchored-preset aware — for anchored presets (e.g.
liangshen/ 梁神) the requirement is skipped during the first pre-promotion round, then automatically applied after promotion. - Clean toggle-off — turning the switch off removes the section from subsequent assemblies; it never rewrites history or inserts a visible chat message.
Installation
From the plugin market (once listed), or:
dsh plugin --profile web add github:dawnliming/dsh-chinese-mode
Restart dsh web, then flip the 中 switch next to the permission selector in the input box.
Settings
| Field | Default | Description |
|---|---|---|
enabled |
false |
Master switch (the 中 toggle). |
text |
(Chinese requirement) | The exact prompt text injected into the system prompt. |
anchoredPresetKeywords |
liangshen, 梁神, anchored, 锚定 |
Preset-name keywords treated as anchored mode. |
skipAnchoredFirstRound |
true |
Skip injection during the anchored preset's first (pre-promotion) round. |
force / customKeywords |
— | Kept for backward compatibility; no longer affect injection (injection is unconditional). |
How it works
The host half wraps systemPrompt.assemble and also listens on the official system-prompt/assemble waterfall event. After the assembly is built, it inserts the dsh-chinese-mode:language section just before the persona (or appends it), so the rendered system message sent to the model always carries the Chinese requirement. The client half renders the 中 switch bound to the dsh-chinese-mode settings namespace.
Development
# host side: lib/index.js
# client side: lib/client.js (ModuleLoader bundle, requires rebuild after edits)
To install the local checkout into a profile as a file: dependency:
cd ~/.dsh/profiles/web
pnpm add file:../../plugins/dsh-chinese-mode
# then add "dsh-chinese-mode" to dsh.profile.bundles in package.json (or let dsh reconcile it)
License
MIT
Links
More in this category
feibi-mochi/deepseek-harness-wallet★ 17
Multi-provider wallet chip: official DeepSeek balance, per-session cost & tokens, third-party token totals, recharge shortcut, low-balance alerts.
Mars-Sea/dsh-commandcode-provider★ 14
Unofficial Command Code LLM provider: registers a `commandcode` route with a live model catalog and reasoning-effort support.
franksong2702/dsh-codex-connect★ 11
Connect ChatGPT OAuth and OpenAI Codex models to DeepSeek Harness, with opt-in search and image tools.
jyh20030112/dsh-visual-plugin★ 9
Gives text-only models vision: forwards user images to an OpenAI-compatible vision model and shows the descriptions in a Web UI right panel.
suntianc/dsh-codex-auth★ 7
Reuses the Codex CLI ChatGPT login as an `openai-codex` LLM route and adds GPT Auth controls to DSH Web settings.
Noob-stupid/dsh-github-login★ 6
Visual GitHub login without a terminal: in-window device flow, token synced into the gh CLI, and host status/launch endpoints.